CN1330810A - Touch-typable devices based on ambiguous codes and method to design such devices - Google Patents

Abstract

The design of typable devices (6005), in particular, touch-typable devices (6005) embodying ambiguous codes (6001), presents numerous ergonomic problems. Solutions for these problems are herein disclosed. This invention teaches methods for the selection of ambiguous codes (6001) from the classes of strongly-touch-typable ambigous codes (6001) and substantially optimal ambiguous codes (6001) for touch-typable devices such as computers, telephones, pagers, personal digital assistants, smart cards, television set-top devices and other information appliances, given design constraints such as the size, shape, and computational capacity of the device, the typical uses of the device, and conventional contraints such as respect of alphabetic ordering or Qwerty ordering.

Description

Keyable device designed according to different codes and methods

This invention relates to keying devices and their use in computing and electronic communications. And especially related to keyable devices designed according to easy-touch differential codes and optimal differential codes.

Since the invention of the typewriter 100 years ago, keyboard engineering has been a very active field of research and development, resulting in many competitive designs. With the growth of personal computers and electronic communications, the number of keyboard designs has increased dramatically as designers have to accommodate different kinds of constraints and seize the opportunities presented by this new technology. However, the variety of these prior art keyboard designs is not due to different kinds of constraints and opportunities, rather, it is a result of keyboard designers not having a complete understanding of the existing constraints within the problems they are trying to solve. At the same time, it also reflects the lack of a holistic, efficient approach to optimizing these constraints. Therefore, the state of the art at this stage has a plethora of incomplete solutions. But these shortcomings can be overcome by the keyboard design method of this invention fully. To demonstrate this invention, the optimization method will be applied to the design of multiple device realizations, each of which is the best basic solution under a series of design constraints.

The present invention relates to keyable devices. Touch typing, like playing an instrument, is a fairly difficult hand skill to learn. Once learned, it is difficult to correct established movement patterns. This places considerable constraints on keyboard design. The well-known Qwerty keyboard (and its close variants such as the Azerty keyboard commonly used in France) has a great advantage because of its deep-rooted movement patterns and over-learning. Therefore, the large range of the Qwerty keyboard makes learning more advanced keyboards, such as the Dvorak keyboard, difficult. Although the Dvorak keyboard also has its users, but the number is not many. However, the Qwerty keyboard is not suitable for palm-sized or small typing devices because of its large number of keys. The advent of palm-sized or small typing devices has brought a silver lining to keyboard designers. Because the existing repetitive movement pattern tends to be fixed, if the new design that appears in this opportunity has a dominant position, it will continue to maintain its dominant position in the future, even if new designs continue to appear in this field . This places a huge burden of responsibility on keyboard designers to avoid suboptimal designs for future keyboard users.

In this prior art, by editing a series of symbols, there are two main methods to reduce the number of keying modes 1) coordination method, combinations of keying modes edit each symbol, 2) divergent coding, each keying mode editing combination of symbols. Coordination is not practical because learning to do it is difficult and few people are willing to invest the time it takes. Therefore, only divergent coding, or a combination of divergent coding and harmonization, can give a real solution to this problem.

The purpose of this invention is to provide a method for the design of the best divergent code in the typing device.

The further purpose of this invention is to provide a method for the design of the easy-touch key type differential coding in the typing device.

The further object of this invention is to provide a suitable keyboard for touch typing of ordinary and small keyboards.

A further aim of this invention is to allow alphanumeric information to be transmitted from a normal telephone or two-way pager to another device of the same type without human intervention, making the service cheaper.

A further aim of this invention is to provide touch-key personal digital assistants.

A further object of this invention is to provide the driver of a car with a keyboard that can type without and without distraction from driving.

A further object of this invention is to provide a key-in communication device which is inexpensive for manufacturers and which can be combined with a general telephone communication system.

The purpose of some preferred embodiments of this invention is to help typists trained with traditional keyboards transfer their typing skills to adapt to the new keyboard by retaining some of the designs that traditional keyboards have on the new keyboard.

A further object of this invention is to provide an overall method for producing divergent codes with the lowest search error rate.

A further object of this invention is to provide an overall method to produce discrepant codes with the lowest query rate.

A further object of this invention is to provide a device that can reduce typing injuries.

A further object of this invention is to provide a secondary foldable palm calculator device.

A further object of the invention is to provide a one-handed keypad suitable for being mounted on a palm calculator.

A further object of the invention is to provide a one-handed and two-handed keypad suitable for installation on a palm calculator or a desktop keypad.

A further object of this invention is to provide an easy-to-learn coordinated keyboard.

A further object of this invention is to provide a synergistic blend of harmonious and divergent keyboards.

The further purpose of this invention is to provide a touch-type input-oriented query mechanism for input devices with different codes.

A further object of this invention is to provide a touch-key-oriented disambiguation mode for keyable devices programmed with disambiguation codes.

It is a further object of this invention to provide a hybrid harmonized/differential coded keypad which can be fully integrated with a standard telephone system.

A further object of this invention is to provide ergonomic designation symbols for patterns. A further purpose of this invention is to provide a transparent touch key interface for a typed device with a touch screen.

A further purpose of this invention is to provide optimization among a set of natural languages.

A further object of this invention is to provide a keying device that can be held with one hand and has a reduced scanning time.

Other purposes of this invention will be described in detail in the following paragraphs.

A short description of the pattern

The preferred embodiment of this invention will now be discussed in detail with respect to the drawings, the following of which are short descriptions.

Figure 1 shows an overview of the optimization considerations for a keyable device made according to this invention.

Figure 2 shows a flow chart of the construction of a device manufactured according to the touch-sensitive differential code.

Figure 3 shows a flowchart for the construction of a divergent code that meets at least one ergonomic criterion and is optimized with respect to these ergonomic criteria.

Figure 4 shows a flowchart for a specific implementation of the method of Figure 3 using the on-demand optimization method.

Figure 5 shows the distribution of probability of finding errors for some of the on-demand discriminative codes on selected keys.

Figure 6 shows the distribution of query likelihoods for some on-demand differential codes on selected buttons.

Figure 7 shows the flowchart of guided stochastic forward optimization.

Figure 8 shows the flow chart of the construction of the easy-touch differential coding.

Figure 9 plots the lookup error rate versus the number of on-demand buttons, and optimizes the divergence code.

Figure 10 plots the query rate versus the number of on-demand buttons, and optimizes the divergence code.

Figure 11 shows the lookup error rate versus lookup rate for optimally divergent codes over a range of button numbers.

Figure 12 shows a graph related to the level of accessibility of the number of buttons to achieve this level, under several different optimization methods.

FIG. 13 shows a flow chart of the method for synthesizing encoding symbols.

To help the reader understand the unity of the device incarnations of this invention, the table in Figure 14 provides a summary of these incarnations and their characteristics. These particularizations must clearly and unambiguously point out the broad scope and different aspects of the invention.

Figure 15 shows an embodiment of a smart card with 16 keys for encoding alphabetic symbols.

Figure 16 shows an embodiment of a smart card with 9 keys for encoding alphabetic symbols.

Figure 17 shows the keyboard embedded in the steering wheel.

Figure 18 shows that 10 keys in the phone have optimized codes.

Figure 19 shows the reduced-disparity alphabetical ordered-disparity code applied to a mobile phone.

Figure 20 shows a Qwerty-like keyboard with optimized lookup error rate and query rate, and involves the arrangement of letters in each row of the Qwerty keyboard.

Fig. 21 shows another keyboard arranged like Qwerty.

Figure 22 shows a divergent keyboard with a standard number pad arrangement.

Figure 23 shows a disambiguation mechanism with ergonomic touch-sensitive guidance.

Figure XXIV shows a flow chart of the method. This method answers queries in a touch-tap guided manner.

Figure 25 shows the design of a one-handed keyboard. This design preserves typing skills between one-handed and two-handed keyboards.

Figure 26 shows the design of a two-hand keyboard. This design preserves typing skills between one-handed and two-handed keyboards. In this case, the two-handed keyboard is selected because of the greatest similarity in the typing state among the two keyboards.

Figure 27 shows the design of a two-hand keyboard. This design preserves typing skills between one-handed and two-handed keyboards. In this case, the two-handed keyboard can be equally distributed between the two hands.

Figure 28 shows a keyboard combined with a mouse.

Figure 29 shows the front side of a foldable information device in an unfolded state.

FIG. 30 shows the reverse side of a twice-foldable information device in an unfolded state.

Fig. 31 shows a double-foldable information device in a state of being folded once, showing its additional functions.

FIG. 32 shows a double-foldable information device in a state of being folded twice, showing another function.

Figure 33 shows a double-foldable information device in a separated state, which can be typed with both hands.

Figure 34 shows a typical PDA with a touch screen.

Figure 35 shows a typical PDA with a potentially transparent keyboard.

Figures 36A, B and C show three modes of a 16-key keyboard.

Figure 37 shows a standard telephone arrangement.

Fig. 38 shows a hybrid keypad of harmonized/divergent codes programmed into the phone.

Figure 39 shows the distribution of lookup error rates and lookup rates for a mix of all harmonized/divergent codes of a particular structure compared to the lookup and lookup rates of a standard divergent code.

Figure 40 shows the flow chart of the construction of multi-level easy-touch differential codes.

Figure 41 shows a flow chart of the construction of a specific embodiment of a multi-touch differential code.

Figure 42 shows a keyable device suitable for the multi-level differential encoding of Figure 41.

Figure 43 shows the operation of the apparatus of Figure 42, showing the first level of multi-level divergent encoding.

Figure 44 shows the second-level encoding of the multi-level divergent encoding.

Fig. 45 shows the operation of the apparatus of Fig. 42, showing part of the second stage of multi-level divergent coding.

Figure 46 shows the sequence of operating states when the device of Figure 42 is keyed into "think" using a multi-level divergent code.

Figure 47, like Figure 46, shows the sequence of operating states when the device of Figure 42 uses a multi-level divergence code to key in "think". But in this case, the operation of the visual cache to reduce the scan time is also displayed.

Detailed description of the invention

Definition and Basic Concepts

This part is the definition of characters and concepts, which will appear in the following detailed description.

A language is a set of symbols in which an individual can establish an order and determine the probability of the order. Groups of signs, sequences of signs, and the possibility of these orders are here language. For clarity of discussion and without limiting the scope of this invention, we refer to language as written natural language, such as English, although specifically we may refer to symbols as "letters" or "punctuation", but as far as this technology is concerned In terms of general technique, the symbols discussed here may also be any unrelated writing units, including standard symbols, such as Chinese ideographic characters, or symbols created by the singer "Wang Zi" after changing his name.

Keyboard/Keying Mode The keyboard is a part of the communication and/or computing device. The keyboard translates the operator's physical movements into a sequence of symbols. A keyboard consists of at least one keying modality responsible for translating a subset of the physical actions of the operator to activate a subset of the symbol sequences of the keyboard.

The physical movement of operating the keyboard is usually the movement of fingers and/or thumbs or a palm-sized stylus. This definition also extends to other bodily movements, such as movement of the head, tongue, or eyes. These movements may be used to signal symbol selection from the keyboard. A device with this keyboard definition is a typeable device.

We understand that a "typeable device" is not only a device with a keyboard, but also an entire communication system embedded in this typeable keyboard. The limits of this system are defined by the basic divergent coding framework. If the keyed symbols in a keyable device appear directly on the display screen, and the display screen is part of the keyable device, then the limits of the system are quite clearly defined by the size of the device. In the more common case, a keying device contains a telephone handset which transmits information to a central computer which then takes care of encoding or processing the original information from the handset. The keyable device of course includes the central calculator, and its mode of operation is set by software built according to the teachings of this invention.

At least one way of typing within each keyboard requires a great variety of physical expressions. One key feature is that it allows an operator to select a subset of a set of symbols to be encoded by the keyboard. With this understanding, and to increase the readability of this manual, "key" is often interchanged with "type".

Typing is the process of sequentially selecting at least one type of typing for the purpose of selecting a subset sequence of symbols from a set of symbols encoded by the keyboard. Well-known handwriting recognition software allows typing to translate a series of drawing actions into a series of subsets of a set of symbols.

Touch typing The process by which a sequence of symbols is produced by the keyboard, most or only using kinesthetic feedback rather than visual or auditory feedback.

Extremely correlated symbols and symbol sequences are well known, and different letters appear at different frequencies in characters. For example, in the previous sentence, the letter "e" occurs 11 times, while the letter "z" does not occur once. This also happens with two-letter words, three-letter words, and so on. Quite clearly, all characters do not appear at the same frequency. The three-letter word "the" is common in English, while the three-letter word "zap" is quite uncommon. These statistical irregularities can be used in the design of divergent codes. Indeed, statistical irregularities have been commonly used in keyboard design at least since the invention of the Qwerty keyboard.

We are also particularly concerned about certain symbols and sequences of symbols whose distribution in a typical sample of textual content is strongly correlated with other symbols or sequences of symbols. Such symbols are known as extremely correlated symbols. For example, the symbol "." is used in English and other languages to signify the end of a sentence, and this symbol may be extremely relevant because the distribution of sentence lengths within typical textual content is not random. In Hebrew, the symbol "." is also associated with certain alphabetic symbols, although Hebrew uses different symbols at the end of letters, and the end of a sentence is also associated with the end of a character.

Reference data A sequence of symbols used to measure inter-symbol correlation, the reference data being generally estimated by analysis of a reference corpus. A corpus is a fairly large collection of textual content, selected to represent some part of a language. Linguists are well aware that there are many fundamental problems with building a corpus that is representative of a language as a whole, sometimes blocking features associated with a particular kind of text or a particular kind of author. These problems are beyond the scope of this invention. Here we take reference data collected from the British National Corpus throughout. The British National Corpus is the largest corpus of analytical English in existence. Selecting a corpus is an essential step in gathering results, allowing methods and specificities to be compared. The establishment of this selection should not be used to limit the scope of this invention. In particular, the choice of English text content and database is an arbitrary choice. The same analysis can also be used for any other written natural language.

Coding and decoding In the United States, the keys on the telephone keypad are usually marked with letters and numbers at the same time, the key with the number 2 is also marked with a, b, and c; the key with the number 3 is also marked with d , e and f, continue in alphabetical order.

Therefore, press the key sequence of 233 and also press out alphabetical order add, bee and bed simultaneously, all are English words, certainly also produce other meaningless alphabetical order as cff. An order is considered meaningful if it appears in a meaningful order reference list. So these alphabetic sequences, whether meaningful or not, are related to the same numerical sequences. We call the key button sequence by 233 coding, and the order of add, bee, bed, eff, etc. is the decoding of coding 233. To clear up the confusion, "decoding" means "meaningful decoding". Taking this as an example, depending on the alphabetical order, the group of symbols used for decoding are the decoding symbols, or just the symbols, without any confusion. And the group of symbols used for decoding, in this case, numbers, are considered to be decoding symbols.

Divergent Encoding Divergent encoding is well known in the art. On the standard telephone keypad in the United States, there are 12 keys in total, 10 of which are encoded with a number, and several of them are encoded with 3 or 4 letters, arranged in alphabetical order. These assignments produce a divergent encoding, which we call the standard divergent encoding. This is encoded as abc def ghi jkl mno pqrstuv wxyz.

Since each key is programmed with several letters, a method of disambiguation must be used to determine which letter operator wants to use. In common applications, such as voice messaging systems, the determination of which letters to use is made by comparing the keyed sequence with a stored reply list. If several replies in the stored replies list match the keying sequence, the operator must choose by himself. The order in which these responses are displayed may be a matter of personal decision, or depending on the frequency of which responses are the correct responses, and the responses are presented in the decoding order of the frequencies.

Standard keyboards There are three main types of standard keyboards that are widely used: the Qwerty keyboard and its close variants, the 12-key telephone keypad, and the 17-key numeric keypad and its close variants. This invention has the unique advantage of providing a useful method for standard telephones, numeric keypads, and specially designed keypads not covered by this description.

Accessible keys are non-accessible typing devices, that is, the symbol designation of the key buttons is fixed; only a typist familiar with the device can develop the physical instinct to use specific movement patterns to decode specific symbols. Devices with easy touch keys have the following three characteristics: 1) low touch keys, 2) based on differential coding, and 3) in normal operation mode, touch typists can use typed devices to type out text with a certain acceptable level. Accurate text content without the need for undue distraction from the work of touch-typing to interfere with the disambiguation process.

Touchability is a matter of degree; it is a measure of touchability. Tactileness depends on many factors, some of which are personal to the typist, some of the use of the typing device, and some of the structure of the typing device itself. For example, a typeable device may be accessible enough for one touch typist for some typing tasks, but not for others.

The two key definitions of touchability, textual precision, and touch-typist distraction depend on a number of factors, including:

□ method of disambiguation,

□ The circumstances in which the machine is used, such as when driving a car or sitting at a desk,

□The type of text content that needs to be typed, this text content determines the necessary degree of accuracy,

□ Reference data,

□ typist's skill,

□ personal preferences, and

□ The way the disambiguation mechanism engages the user (eg, when the speech synthesis mechanism speaks a sentence or question, the mechanism may or may not be more distracting to the user than a bell or flashing light).

While accessibility, like temperature, is a matter of degree, it, like temperature, is very clearly defined. Once these various factors have been determined using standard experimental protocols well known to those skilled in the art, the accessibility of a keyable device, in relation to the user or finger group used, can be quantified. More specifically, two components of accessibility can be measured in terms of discrepancy codes: bug-finding and querying. Therefore, the value of the ease of touch can be specified without directly referring to any number of users, but only by referring to the aforementioned divergent codes.

Like temperature, there is a lower bound on accessibility. We clearly know that for any typist, a device that requires user intervention to resolve the ambiguity after every character or every three characters is typed cannot be said to be an easy-touch device. The lower boundary of touchability can be shown by the persistence of attention. If a typist's attention must always be focused on the operation of the disambiguation mechanism to type acceptable text content, then this device is not easy to touch.

The actual lower bounds for touchability are associated with typists with average typing skills, and are higher than the theoretical lower bounds above. In order to make the inventive concept of touchability precise and accurate both numerically and conceptually, the numerical value of touchability is presented as the value of finding errors and queries. This numerical characterization more clearly indicates that the methods and apparatus of the present invention differ from those of the past.

Accessibility differential coding is the basis of the Accessibility Device.

Feedback device This device allows the user to intervene at different points to decode the sequence of symbols produced by the divergent code, within which feedback to the user's sensory perception is necessary. Usually this kind of feedback is presented in the form of symbols, but there are also many forms of feedback, such as auditory, tactile and even olfactory.

Ergonomic factors relate to the design of keyboards containing ambiguous codes and the elimination of constraints. This may include a reduction in search error rates, a reduction in lookup rates, the number of keys that fit the size of the keyboard, the degree of integration with existing keyboards such as Qwerty keyboards, telephone keypads, or numeric keypads, the stability of the divider structure, Structural precision, minimal use of mode-shifting buttons, divider construction, integration between one-handed and two-handed typing, and preservation of traditions such as alphabetical order. Other constraints include: the ergonomics of the disambiguation mechanism, the ergonomics of decoding symbols of low relative efficiency, the look and feel, and the availability of computer resources at the transmitting and receiving ends of communications utilizing disambiguating codes.

Lookup errors measure errors produced by a disambiguation mechanism that systematically selects the most likely (most meaningful) decoding from a sequence of possible decodings in a disambiguating order to disambiguate. Thus, the search error rate for a code is the total number of possible decodings divided by all possible decodings that are not the most probable decoding of the divergent order, which is the reference likelihood of possible decodings. In the case of character-based disambiguation, these sequences begin and end with "spacebar". That is, these sequences are characters. Lookup errors refer to the likelihood that the most likely decoding is not correct. Lookup errors are presented as a ratio. The lookup error rate is the ratio of characters to lookup errors. The lookup error rate is the inverse of the lookup error probability.

The query likelihood is the total divided by all non-unique (meaningful) decodings which are the reference likelihoods of the preceding decodings. This points out the possibility that a character may have more than one meaningful decoding. Therefore, the user must use a query to determine which decoding to use. The reciprocal of the query likelihood is the query rate, the unit is the ratio of characters to queries. Query rate provides the average number of characters typed between queries.

Substantial optimization A code is said to be substantially optimal if it is one of the best codes in terms of performance, taking into account other constraints imposed on the code. For example, the search error rate of the coding on the 20 key buttons may be lower than that of the coding on the 2 key buttons, but considering the limitation of the coding on the 2 key buttons, the coding on the 2 key buttons may be It is optimized for the search error rate. , an optimal ergonomic code can be defined as the code that is optimized for each set of ergonomic constraints simultaneously.

These limitations include but are not limited to the number of keys, search error rate and query rate. Of these three constraints, every pair of constraints is related. Lookup error rates tend to increase along with lookup rates, and both lookup error rates and lookup rates increase as the number of keys decreases. When a known criterion is the only optimum criterion and some other criterion must also be optimized, the best possible value of the known criterion may be better than the best possible value available. Therefore, the ergonomic constraints associated with known designs and their importance must be seen as an initial step in the optimization method proposed by this invention.

It must be emphasized that there is no way to definitively discuss the optimization of divergent encodings, but one must evaluate the correlation with a reference data set for the language to be encoded. Indeed, it is possible to create data sets and code them optimally, taking into account any discrepancies in coding, with regard to the data already created.

Considering the reference data set, an estimate of the optimization of a known code can be obtained experimentally with random codes. Random encoding will be discussed in detail later.

Disambiguation Methods To properly define the optimality of a disambiguation encoding substance, one must consider the disambiguation method that has been chosen. An optimization associated with one disambiguation method may not encode an optimization for another disambiguation method.

Among the techniques, there are at least two well-known disambiguation methods. These are character-based and block-based disambiguation methods. In character-based disambiguation, a list of characters and their possibilities are used to select between alternative decodings of known codes in a divergent code. For example, all characters in a known encoding that have meaningful codes will be compared, and the word with the highest probability will be selected. The block-based disambiguation method is quite similar, the difference is that the list contains fragments of text content and the possibility of fragments, and the scale of these text content fragments has a certain degree.

Both character-based and block-based disambiguation are special cases of the general framework. The general framework is an order-based disambiguation method, in which the ordered list is related to the possibility, and the disambiguation is performed by this list. It is important to note that in languages such as English, the "whitespace" notation that defines word boundaries is the purpose of this discussion, no different than any other notation. One can define a list of sequences and sequence capabilities in which the aforementioned sequences include "blank" symbols, thus extending beyond word boundaries. Individuals can further define sequences including wildcards, and then define sequence lists containing arbitrary sequences that do not necessarily correspond to literals within the language. In such cases, arbitrary compound representations of the language can be built and used in disambiguation.

For example, syntactic and semantic relationships between subsequences can be used to resolve conflicts between possible interpretations of ambiguous coding sequences. To make it easier to understand, in this application, unless otherwise specified, we will use the well-known character-based disambiguation method as the main method. Those skilled in this method will be able to appreciate the benefits of the method advocated by this invention in that it is not necessary to use the character-based disambiguation method, and any other disambiguation method is available.

A division of an integer n is a division of integers whose sums add up to n. Usually a known integer has several divisions, for example, the integer 5 can be divided into 3:2 or 2:2:1. Those skilled in this method should be familiar with the division method used to generate all possible divisions of an integer. Most traditional codes use the even-as-possible-separation method. That is to say, in a partition, the number of designated keys is obtained according to the number of letters of the required coding within the possible range, and the number of letters on each key is the same. It will be explained in more detail below that this choice is a reasonable choice for some ergonomic considerations, although it may be a secondary consideration in others.

There are two kinds of plurals of divergent codes to apply for their exclusive right here. These two pluralities are 1) easy-touch differential codes, and 2) substantially optimal differential codes. Distinct codes may be substantially optimal but not accessible, or accessible but not substantially optimal, neither substantially optimal nor accessible, or both substantially optimal and accessible.

The discovery begins by showing how to find the divergent codes in these two complex numbers, and to find out in which complex number the code is. What follows is an explanation of how these codes in these two plurals are used to make typing devices, and how these codes are used to solve different design problems encountered by designers who make typing devices.

The best mode of the invention is based on a combination of design criteria which should be optimized in accordance with the guiding principles of the invention. The series of methods and apparatuses advocated by this invention will now be illustrated with a few practically relevant and practical special cases.

The range of devices that can be constructed by those familiar with the method following the principles of the invention goes far beyond the few applications selected here. Many different extreme or special conditions that are design constraints can be resolved in this application. Given the guidelines contained in these situations, it will be apparent to those familiar with the methodology how to properly combine these features in order to address over-neutralization or synthetic design problems.

One such application is optimization in terms of lookup error rate alone. This application is designed for a device with low memory and low calculator capabilities, an example of which is a smart card. In such a device, computer resources may not be available to support a compound query procedure for the user in the disambiguation process. The device therefore uses one of the simplest disambiguation procedures, namely the systematic selection of the most probable decoding for any given code.

Another application is to optimize with reference to query rate alone. This application is designed for drivers of vehicles, an example of which is a car. Even if computer resources are sufficient to support a complex inquiry procedure, the use of such a procedure should be minimized so as to minimize the driver's distraction from driving.

The next application is to provide keypads for telephone sets to be optimized with reference to the search error rate and query rate, while being compatible with the design of keypads for general telephone sets.

Another application is to optimize according to the usual standard: to preserve the English alphabetical order. The letters on the keypad of a general touch-type telephone are arranged in order. By optimizing the separation, it is possible to preserve the English alphabetical order on the keypad of a traditional telephone, and at the same time reduce the search error rate and query rate.

Optimizing the separation provides more of the substantially optimal search and error rate keypads shown, while preserving the arrangement of traditional Qwerty keypads as much as possible.

A further application that illustrates a keyboard design that is as consistent as possible with conventional designs is a keyboard based on diverging codes and consistent with a numeric keypad.

In many applications, an ergonomic keyboard capable of both disambiguating and disambiguating operations is advantageous. By far, it would be preferable to choose a divergent code among some key numbers that divide the desired code sign. A symbolic number key number equal to 1/2 is most commonly used. The ideal result of this common choice will be shown in the next application.

Another application is how the keyboard is optimized for cross-platform compatibility. In this application there are two keyboards, a one-handed keyboard and a two-handed keyboard, which are designed to be operated at high speed transitions so that the touch-typing motions used to operate one of the keyboards are without difficulty Transition to touch-typing actions for operating other keyboards. This keyboard has the added advantage of reducing the possibility of typing injuries, among other purposes and advantages, which will be explained in the detailed description.

Combining all of the above applications shows that different keyboards represent different degrees of optimization, and since a given user may need to use the keyboard in several different situations, it also represents the need to use the keyboard in several different situations. Programs and different solutions coexist on a single device. There is an unexpected solution to this problem due to the size of the small typing device that is only achievable with the help of differential coding, namely the bi-fold personal digital assistant referred to in this application.

The few applications discussed so far are associated with hardware and software specifications. However, it is possible to achieve many of the goals of this invention using an entirely or primarily software-based solution. An example of a software process will be described in detail to explain how the use of suitable software can combine the characteristics of existing hardware to achieve some of the objectives of the invention.

The final set of applications harmoniously combines, for the first time in history, two alternative methods for making low-key keyboards: the harmonized method and the divergent coding method.

First, this demonstrates that an ergonomically built coordination model combined with an optimization program for look-up rates and look-up error rates will allow for the divergent coding function on an n-key to be substantially larger than an n-key The divergence codes on the m-keys function similarly. When this method is applied on the general divergence code, these 8 letter keys on the general divergence code then obtain the property of the best coding in essence without needing to coordinate on 13 key buttons. All of this discussion of Creation Notes on how to extend divergent encodings, although referenced in English, applies to other languages as well. Specifically, the discussion here is in terms of this application, but the annotations are generally applicable to all applications.

Second, this demonstrates that the number of keying devices can be further reduced by combining a split-occupation method with the exemplary method in the previous application, in this example 4 keying devices are used to operate a 16-code differential encoding device. The number 4 is chosen so that a palm-sized device using this code can be operated via the index finger and thumb of the hand holding the device.

An overview of the operation of the easy-to-key device.

Figure 2 illustrates an outline of the operation of an easy-to-key device based on differential codes. Such a device has a keying function, and operation of the device (step 140) produces a sequence of coded symbols (step 141). Touch-key differential coding is used in step 142 to align these coded symbol sequences with decoded symbol sequences. These coded symbol sequences are then optionally output to a display for direct viewing by the user of the device, or electronically for further processing, transmission or storage (step 142).

It should be particularly pointed out here that the differential coding device in the summary of FIG. 2 also meets other ergonomic criteria besides easy-to-touch keys.

Establishment of the Substantially Optimal Code The steps in the research method for optimizing a divergent code with reference to a set of ergonomic criteria are illustrated in FIG. 3 . Overall, the steps are as follows:

□ Step 2000 selects a set of statistically correlated decoding symbols to implant into a divergent code, which includes the following sub-steps

2007 select a set of reference data,

2008 According to the statistical correlation of the data analysis symbols selected in step 2007 Step 2001 selects a disambiguation method.

□Step 2002 selects coded symbol numbers.

□Step 2003 to select the ergonomic standard according to the code which should be the best in nature.

□Step 2004 compares the importance of the ergonomic criteria selected in step 2003.

□Step 2005 to select an optimization method.

□Step 2006 applies the optimization method selected in step 2005 and produces

Get the substantive best divergence code.

It can thus be seen that steps 2000 to 2003 can be applied in any order, and the application of one of these steps will be able to influence the selection in the other steps. Details of the application of these steps will now be described.

Step 2000 selects a set of statistically correlated decoding symbols to implant into a divergent code. This step also includes the sub-step 2007 of selecting a set of reference data, and the sub-step 2008 of analyzing the statistical correlation of symbols according to the data selected in step 2007. The purpose of these steps is to find out which symbols can be differentially implanted. All disambiguation procedures use the correlation between symbols to predict which sequence of decoded symbols should be concatenated with a sequence of encoded symbols. If a decoding symbol is randomly distributed in the text to be encoded, it cannot be embedded in a divergent code, because it is impossible to predict a randomly distributed symbol. In general, for any natural language, the symbols used to encode that language (eg, letters in English and ideograms in Chinese) will be statistically sufficiently correlated to facilitate the design of efficient differential encoding of these symbols. There are also other symbols, such as punctuation marks, that may be statistically sufficiently correlated between these symbols and with the letters or ideograms used to write the language. The minute details of steps 2007 and 2008 will depend on the language to be represented. Methods for analyzing the statistical correlation of symbols used in natural language scripts are quite familiar to linguists.

Step 2001, choose a disambiguation method. As mentioned above, there are at least two well-known disambiguation programs, block-based disambiguation programs and character-based disambiguation programs. Both methods use statistical correlations between symbols to predict which sequence of decoded symbols should be concatenated with a sequence of encoded symbols. Both block-based and character-based approaches can be augmented by using higher-level information about the language, such as its syntax and semantics. The purpose of these steps is to create an ambiguous code subject to the chosen disambiguation procedure such that the most selective linking of the decoding sequence to each coding sequence is made. The details of the disambiguation procedure chosen can affect the detailed nature of the disambiguation codes designed in this way. This method will choose the character-based disambiguation procedure as the disambiguation procedure to illustrate, although other disambiguation procedures will also be discussed.

Step 2002, select coded symbol numbers. The selection of code symbol number is extremely important for the design of the typing device based on differential coding. This selection was made considering a number of factors, including the size of the typing device and the acceptable degree of ambiguity. These factors and their interplay are best explained with reference to practical examples, which are used below.

In step 2003, the ergonomics standard is selected according to the code that should be substantially the best. A focus of this invention is to discover and define several divergent codes of ergonomic criteria that determine the quality of a typing device. These criteria include accessibility, lookup error rate, lookup rate, structure accuracy, entity accuracy, preservation of legacy, compartmentalized structure, cross-platform compatibility, design regularity, and scan speed. Depending on the individual application, more than one standard may be applicable to the design of the typing device.

Step 2004, comparing the importance of the ergonomic criteria selected in step 2003. While more than one criterion may be applicable to the design of a typing device, some decisions must be made on these criteria in terms of their importance. It rarely occurs that the same optimum value of a known ergonomic criterion can be achieved when this ergonomic criterion is considered in isolation and when at the same time it is necessary to optimize another ergonomic criterion.

Step 2005, selecting an optimization method. Two optimization methods, random selection and guided random progression, are discussed in detail below. Random selection between the two is usually easier to apply, while guided random progression produces better codes. These two optimization methods are representative of many that may be applicable to a given typing device design. In some cases, such as the first concordant/dissimilar coded installation considered below, the number of codes to examine is small enough to make a thorough comparison of each.

In step 2006, the optimization method selected in step 2005 is applied, and thus a substantially optimal discriminative code is produced. Regardless of the optimization method chosen in step 2005, certain techniques must be used in applying the method that produces the substantially optimal disparity code. Especially when an optimum value is immediately required for several ergonomic criteria, it is better to first consider each ergonomic criterion individually, so that a prediction can be made about the final achievable encoding quality. This prediction can be considered invaluable for fine-tuning optimization, which is discussed in more detail below, once the two optimization methods are fully introduced.

Random Selection The basic method of finding a code with good properties is to randomly select codes, test the properties of the codes, and choose the code with the best properties. Exhaustive enumeration, ie testing all codes in the set of candidate codes, is usually not a viable option since the number of codes to be tested is too large for any reasonable amount of computation time.

Random selection provides a benchmark against which to measure the performance of other code selection methods. Assume that a set of ergonomic criteria and a comparison of this criteria are known. We can predict the substantive optimality of a first divergent code with reference to this criterion and its comparison by randomly selecting additional divergent codes. A first code is not substantially optimal if it is possible in a small number of random tests to find a code with equal or better values than the first code with reference to known ergonomic criteria.

Conversely, if it can be shown that a substantially greater amount of random testing is necessary to produce a code having a better or equal value to the first code, or if a better code exists, then the first code That is, not substantially optimal.

With reference to Figure 4 we detail a method for ruling out the assumption that the candidate ambiguity codes are substantially optimal. Overall, the steps are as follows:

□ Step 3000 determines a set of relative constraints that define a suitable encoding group containing candidate codes.

□Step 3001 to determine a set of ergonomic criteria in which the candidate code may be substantially the best.

□Step 3002 Randomly select a coded subset from the code group determined in step 3001.

□Step 3003 evaluates the individual codes selected in step 3002 according to the ergonomic criteria determined in step 3001.

□Step 3004 compares the value of the candidate code according to the ergonomic standard determined in step 3001 and the value obtained in step 3003. If any value obtained in step 3003 is better than the value of the candidate code, the assumption that the candidate code is substantially the best can be ruled out. The details of these steps are as follows:

In step 3000, a set of relative constraints is determined that defines a suitable encoding group that includes candidate codes. In case a candidate code needs to be evaluated for its substantial optimality the code group must accept the appropriate definition. Relevant constraints to some of the possibilities are: the number of coded symbols, the separation structure and the recognition of a specific arrangement, such as alphabetical order. Each constraint places a limit on the set of codes that have been properly compared with the candidate codes.

Step 3001, determine a set of ergonomic criteria in the case where the candidate code is likely to be substantially the best. Some of the criteria that may be relevant to the analysis of candidate codes are: error-finding error rate, look-up rate, approval of a specific arrangement, such as alphabetical order, approval of a conventional design, and structural accuracy.

Once steps 3000 and 3001 are performed, the distribution of the coding property of the code group will be defined, and the values of the distribution can be randomly sampled. Taking FIG. 5 as an example, the block codes can be defined as 1) as equal separation as possible and 2) with 7, 9, 11, and 13 as the specified number of symbol codes. The above is the definition of step 3000 . Completion of step 3001 confirms that the search error rate is the only relevant ergonomic criterion. Steps 3002 and 3003 combine the above steps to obtain a numerical distribution, and the graph of this distribution can be obtained by random sampling. Taking Fig. 5 as an example, select 5000 codes from each distribution according to step 3002, and then measure the search error of each code according to step 3003. This value is expressed as the ratio of the lookup error percentage (reciprocal of the lookup error rate) to the number of codes with a known error percentage. As the number of keys increases, the peak of the distribution graph becomes more obvious. If the above steps are repeated with query possibility instead of search possibility, the data in Figure 6 can be obtained.

To illustrate step 3004, a candidate code must be selected to be tested. This encoding is the 14-key encoding such as pn gt cr zk wj a e hi so ud xf ym vl qb proposed in [1]. The comparison between the lookup error of this code and the reference data is that there is one lookup error every 105 words. Following the steps above, we found that 14-key codes separated by as equal a distance as possible could be obtained in an average of seven random tests with a lookup error equal to or better than that of the candidate codes. By repeating the above steps with query likelihood as the relevant ergonomic criterion instead of lookup likelihood, we found that codes with better than candidate code query rates (one query per 4 words) appeared on average three out of every four in random testing . Thus [1] known ambiguous encodings are substantially optimal both in terms of lookup error rate and lookup rate. Indeed, most codes are better than this known code in terms of query rate.

In general, if a code is not optimized according to ergonomic standards, there is a good chance that the code is not substantively optimized with appropriate linguistic data.

guided random progression

Guided random progression is an iterative optimization method. This method has at each step new codes based on the best codes from the previous step, and these new codes may be better than the best codes from the previous step. As the steps are repeated, better codes are produced. The procedure is briefly explained here before further formal discussion.

In the present case, the continuation of the best direction to find is as blind to anyone as we lack this knowledge regarding the divergent coding optimization of more than one ergonomic criterion. unclear. So the safest way is to take as many small steps as possible in as many directions as possible, and never blindly take other actions before evaluating and comparing all the steps and directions. The accumulation of small steps may cause the seeker to hit a wall. This type of inquiry should have a "restart" step, which can bring the situation out of the situation where it cannot be compromised and blocked.

Formally, the main problem is to take the smallest steps in the space of divergent encodings, and then guide these smallest steps through the space above to reach the ideal encoding. In accordance with the teachings of this invention, the substantially smallest step in the space of divergent codes corresponds to a single pairwise permutation assigned to the codes. In each step of the optimization method, the more pairs of paired permutations can be tested, the better, and it is ideal if every pair of paired permutations can be tested. Finally, the paired permutation with the largest qualitative improvement is selected for optimization, thus ending all steps. If there is no paired permutation that qualitatively improves the most, one of them is chosen at random.

Referring to Figure 7, the steps of this method are as follows:

□4000 Select a group from a series of candidate codes as the starting code.

□4001 It is more appropriate to use the perturbation method to obtain a group of new codes from the initial code, and it is more appropriate to use the paired arrangement of symbols and buttons, and it is best to use all possible paired arrangements.

□4002 Measure the properties of this group of new codes.

□4003 Check whether the termination criterion has been reached, which refers to the criterion that restricts further improvement.

□4004 If the termination criterion has been reached, output the best code in this group.

□4005 If the termination criterion is not met, check whether this series of code groups obtained from the starting code by permutation method contains a group of better codes than this optimal code.

□ 4006 If there is a group of better codes than this best code in this series of code groups, select this group of best codes as the new best code.

□ 4007 Select a new start code. If the answer of step 4005 is yes, select the best code in this series of code groups as the new start code, otherwise a group of new start codes can be arbitrarily selected in this series of code groups. After completion, return to step 4001.

When only one criterion is available for optimization, the selection process of selecting the best code from a series of candidate codes will be as easy as selecting a group of codes with the most suitable value on this criterion from the series of code groups. However, when there are more than one standard available for optimization at the same time, the value groups of these standards only have a partial order. At this time, how to make the most beneficial choice for the optimization process among these value groups will not be easy.

One way to do simultaneous optimization is by performing the optimization on each variable individually. Therefore, when a conflict occurs and the synchronization cannot reach the best state, here is the general situation as an assumption, and it is necessary to establish the trade-off of the importance of each standard. And this relative trade-off is part of the constraints on the design.

Establishment of Accessible Codes To help illustrate how accessible codes are made and used, we will set three strict progressive levels of standardization for ease of access.

□Level A This level of easy-to-touch keys can be represented by an informal and acceptable typist. The characteristics of this type of typist are 1) can type 20 characters per minute and will Encounter interference, that is to say, the average typist will encounter a query rate for every 5 words, 2) an acceptable search error rate of 2%, that is to say, the search error rate for every 50 words is searched once, Or every two and one-half seconds.

□Level B: This level of easy-to-touch keys can be represented by a more formal and acceptable typist. The characteristics of this type of typist are: 1) Can type 20 characters per minute and meet every 30 seconds Interference, that is to say, the average typist will encounter a query rate of one question for every 10 words, 2) an acceptable lookup error rate of one percent, that is, a lookup error rate of one lookup for every 100 words, or Every 5 minutes.

□Level C This easy-to-touch key can be represented by a skilled typist. The characteristics of this type of typist are 1) can type 40 characters per minute and encounter interference every 30 seconds, that is to say On average, every 20 word typists will encounter a query rate of doubt, 2) an acceptable search error rate of 0.5%, that is to say, a search error rate of once every 200 words, or every 5 minutes once.

In Fig. 8, we indicate that the method for establishing an easy-to-key coding involves the following steps:

□5000 determines the acceptable search error rate and the quantitative value of the query rate.

□5001 Select a method for divergent coding optimization.

□5002 Determine the minimum required number of keys, that is, use the determined number of keys and steps.

□ The optimized method selected in step 5001 can achieve the search error rate and query rate determined in step 5000 .

□5003 determines the maximum allowable number of key buttons in known typing device designs.

□5004 determines whether these design standards are compatible. If the number of buttons determined in step 5003 is greater than or equal to the number of buttons determined in step 5002, these design standards can be compatible, otherwise not.

□ 5005 If these design criteria are compatible, as determined in step 5004, apply the selected optimization program in step 5001 to create a suitable TTP discrimination code. If it is not compatible then the program is unsuccessful.

The details of this method are as follows:

Step 5000, determine acceptable search error rate and lookup rate values. This can be carried out by testing a single or a group of typists, or directly by selecting the value of the approved search error rate and query rate in advance, such as selecting one of the standard degrees of easy-to-touch keys according to the aforementioned method.

Step 5001, choose a method for divergent coding optimization. Referring to the above method of establishing the optimal divergent code, two optimization methods have been discussed: random search and guided random advancement. Random search is not as effective as guided random advance, but it may be sufficient if the number of allowed buttons is large enough and the perceived ease of access is low enough. Another weaker approach is to choose a yard in a single random trial, which may be more than enough in some cases. To see this in more quantitative detail, refer to the experimental results discussed in Figures 9, 10 and 11.

In this experiment, divergent codes of 5000 and a distribution as even as possible were randomly selected from each group of divergent coded groups of 2 to 20 keys. Simultaneously, an optimization procedure can be carried out on each key button of 2 to 20 with guided random advance method, and this is carried out under the following three conditions, 1) only carry out optimization to search error rate, 2) Optimizing only the search rate, 3) Optimizing both the search error rate and the search rate at the same time using the specified numerical method. The following statistics can be derived from the calculated lookup error rate and lookup rate values for randomly selected codes: best value, worst value, average value, and median value. All of the above statistics are shown in Figure 9, which shows the lookup error rate, and in Figure 10, which shows the lookup rate, and also applies the results of an optimization procedure in which the lookup error rate and the lookup rate are A single ergonomic standard when optimized. The results of an optimization procedure that simultaneously optimizes the lookup error rate and lookup rate are reported in FIG. 11 . From the above data we can make a decision on which optimization procedure to use. Using the most ergonomic method is probably the most obvious choice, but sometimes a lower standard is more than enough, for example, any single randomly selected code can meet these standards if the specified standards are loose enough. This situation will be discussed further below.

Step 5002, determine the minimum required number of keys, that is, use the determined number of keys and the optimized method selected in step 5001 to achieve the search error rate and query rate determined in step 5000.

Referring to the above experimental results, and the selected level of easy-touch typing, we can create a table to display the minimum number of keys required for these three levels of easy-touch typing, and refer to the aforementioned three optimizations method. Figure 12 shows exactly such a table.

Step 5003, determine the allowable maximum number of keys in the known design of the typing device. Differential encoding will be used most often on small devices, and the number of keys is usually limited by the size of the keys and the size of the entire typing device. In some cases, convention may affect the number of buttons, such as the convention of using 12 buttons on a telephone keypad.

Step 5004, determine whether these design standards are compatible. If the number of buttons determined in step 5003 is greater than or equal to the number of buttons determined in step 5002, these design standards can be compatible, otherwise not.

The number of keys that can be allowed on a typing device may be limited by many factors, and these factors determine the degree of limitation, as will be seen more clearly below in detail describing the application of this device.

Step 5005, if the design criteria are compatible, as determined in step 5004, apply the selected optimization procedure in step 5001 to create a suitable easy touch-typing discrimination code. If it is not compatible, the program fails.

If the procedure is to fail, at least one of the following will occur:

□ Select a more ergonomic optimization method.

□The design of this device is improved to accommodate more keys.

□A lower level of easy-touch typing is acceptable.

□This device is discarded.

Smart Cards on the 9th to 16th Letter Keys A smart card is a credit card-sized device containing a calculator-like processor and memory. Smart cards are currently used in applications such as security management and banking, but there are many other possible uses. This application demonstrates that it is possible to mount a smart card on an easy-touch typing keyboard, thereby greatly extending the range of applications that the device can use. To give a simple example, in the application software of security management and banking industry, the user of the smart card must remember a string of numbers used as a password for using the device when using it. However, with a touch-sensitive smart card, a natural-language, easy-to-remember but long PIN string can replace a difficult-to-remember short numeric PIN. Devices similar to the size of a smart card are applicable under the teachings of this invention, including personal digital assistants manufactured by Franklin Corporation and sold under the REX trademark.

With today's technology, the size of the smart card substantially limits the transmission of information keyed on the card through complex and power-consuming communication components. Therefore, the application of this smart card is to promote a low-cost device for ergonomic and rapid information transmission using a conventional touch dialer and a conventional touch tone generator.

There are 12 keys on most telephones, each of which has its own key-touch sound, that is, pressing each individual key will cause the phone to emit a specific key-touch sound. However, the world-wide dual-tone multiple frequency (DTMF) standard provides 16 key-touch sounds, and the dual-tone multiple-frequency audio generator installed on the general telephone set has the function of sending all 16 key-touch sounds. By utilizing these additional tones, each of the 16 keys can have a unique touch sound and can be used to encode alphanumeric symbol sequences. Without considering other factors, the more the number of keys is, the lower the degree of difference in coding of these keys is. So what this application is promoting is actually using all 16 of these key touches to encode an alphanumeric sequence. In this way, devices for transmitting encoded information with low ambiguity can be produced using ready-to-use and low-cost components.

This application also has the following further purpose

□ Provide a tactile keypad for smart card-sized devices.

□ Provide a method to simulate a set of code symbols larger than the set of code symbols that can actually be produced by a device.

□ Provide an example of finding the error rate as the primary ergonomic criterion in an installation.

□ Provide a combination design of keyboard and video display suitable for smart card-sized devices.

□ Provide a disambiguation device that can operate with a very small amount of computer memory.

□ Provide a system with more than one disambiguation device operable, each applicable to its own range of computer-specific functions. In this case, the first disambiguation device is used to provide feedback to the user at the sending end of the communication, and the second disambiguation device is used at the receiving end of the communication.

We will now discuss in detail how this application achieves these goals.

Easy-Touch Keypad The smart card device is small in size, so only a few actual-size keys can fit on it. If a certain part of the card is to be reserved for a video display, then the part where the keys can be installed is even smaller. A good compromise between the ease of touch required for the physical size of the buttons and the need for a large number of buttons to allow low ambiguity encoding is in the range of 9 to 16 buttons. Figures 16 and 15 show two possible device designs for this range of button counts. The arrangement and availability of the buttons and their relationship to other components of the smart card are discussed in detail below.

Synthesizing encoding symbols using sequence encoding and nonsense decoding Referring to Figure 13, let us divide a set of decoding symbols into two subsets: 1) a central group containing symbols that will be connected to encoding symbols thus producing a A one-to-one relationship to devices, 2) An auxiliary group containing symbols that will be linked to coded symbols thus creating a many-to-one relationship to actual keyed devices. (Step 100) In this way, the method for synthesizing coding symbols will include the following steps in one step:

101 Establishing an initial and potentially divergent code relating a subset of the central group to coded symbols that have a one-to-one relationship with the actual typing device that can be generated via the typing device that actually represents the coded symbol.

102 Find short sequences of coded symbols for which no possible decoding of the coded symbol sequence results in meaningful decoding.

103 Establish a relationship between an auxiliary set of coded codes with the possibility of divergence and the short sequence of coded symbols found in step 102.

As an example, let's concatenate 16 coded symbols with 16 dual-tone multiple frequency (DTMF) signals, so specified that the audio will actually represent these coded symbols. These tones will be marked with (0, 1, 2, 3, 4, 5, 6, 7, 8, 9, *, #, A, B, C, D). We will use the letters [A-Z] as the most central group of symbols and connect them to actual representative coding symbols via the following first set of divergent coding

(0, aw), (1, bi), (2, cx), (3, d), (4, ej), (5, fo), (6, g), (7, hv), (8 , ky), (9, 1), (*.mu), (#, n), (A, pz), (B, qr), (C, s), (D, t),

Here the first part of each pair is an encoding symbol, and the second part of each pair is a decoding symbol concatenated with the encoding symbol. The decoding symbols of the auxiliary group will be a single code group containing the symbol "space". From this we synthesize an encoded symbol representing the decoded symbol "space". The candidate sequence is A8A, which corresponds to the following decoding sequence (pkp pkz zkp zkz pyp pyz zypzyz). None of these coding sequences form any of the characters listed on our meaningful sequence reference sheet, so the coding sequence A8A is suitable to represent a part of the auxiliary group, and we form (A8A, "space") This pair is combined to represent the symbol "space". The symbol "space" could then be associated with a keying device which would cause the audio sequence associated with the A8A to be produced each time the designated keying device is activated. At the receiving end, a decoding device converts the sequence A8A into the symbol "space". It is completely transparent to the user whether the known keying device is associated with a single, physically representative encoding device or a composite encoding device. An arbitrarily large auxiliary set of decoding symbols can be represented by this. It can be seen that if there is a set of reference data and a first divergent code of a central group of coding symbols, a programmer with ordinary skills will be able to easily create software that can automatically generate any desired synthetic coding symbol values .

Provides a disambiguation device capable of operating with a very small amount of computer memory. The limited processing power and memory capacity of this generation of smart cards has resulted in substantial low disambiguation in the disambiguation encoding design, meaning that very few calculator functions on the card are dedicated to the disambiguation device.

For any disambiguating code, most of the disambiguation benefits, on the part of the computer's hardware and software, as well as on the part of the user, occur in the choice of which other decoding should be used for the disambiguating code. Due to the limited calculator function of the smart card, questions about other decoding will be completely ignored. In the case of no need to consider the look-up, each code only needs to store its most likely decoding, because only the most likely decoding can be output by the disambiguation device when receiving an individual code sequence . According to this brief description, we can get a particularly small database, for example, a simple suffix tree type. Since queries are not considered sufficient visual feedback can be provided through the use of a simple and power-efficient display, such as a single-line action sign display associated with a display via a pocket computer or digital watch.

This disambiguation method, which only stores and outputs the most probable decoding sequence, we will call it easy-find disambiguation method. The easy-to-find disambiguation method can only operate effectively if the ambiguity code is sufficiently accessible. In this way, we get a surprising conclusion about the coding of easy-touch keys, that is, the codes of easy-touch keys can effectively operate extremely simplified disambiguation devices.

A standard 16-key full-fledged dissimilarity code used in this invention consists of a code with a lookup error rate of once every 4043 words and a lookup rate of once every 68 words, that is, aw bi cx d ej fo g hv This encoding of ky l mu n pz s t is the encoding applied in the standard design 51 of the 16-letter key type smart card shown in Figure 15. The figure also includes a display 50 designed to decode coded symbols typed via the keyboard and a thumb-activated auxiliary keying device 51 that can be used for many additional symbols and mode switches. This is discussed in more detail in other applications. It is worth noting that, if the display 50 can be placed according to the following conditions, then it is preferable on the operation of the input devices 51 and 52: 1) the letter input device 51 and the thumb-activated auxiliary input device 52 are all arranged in a A comfortable position for operation with one hand (the right hand is taken as an example in this figure), while allowing the largest button size, 2) and under the limitation of the size of the smart card, allowing the display screen to have a width between the thumb and index finger Comfortable and complete viewing angle. Such a unique and specific installation method solves the problem of how to make the easy-touch keypad and a maximum possible display co-exist effectively on a smart card.

Since the sender cannot allow lookups in this application, the encoding is selected via a guided random-forward optimization using the lookup error rate as the only criterion for optimization. It should be noted here that the lookup error rate using this encoding will occur about once in every 16 pages of typed text. This code is therefore suitable for the exact transmission of substantially redundant information, albeit in the absence of interrogation means. Assuming that sacrificing some of the optimality of the lookup error rate at the receiver helps reduce query processing at the sender, an alternative encoding optimized for both lookup and lookup error rate criteria is aw bu cx d ev pz go hv im ky 1 NQ pr s t. This encoding has a lookup error rate of once every 2670 words and a lookup rate of once every 101 words. Choosing an encoding optimized in terms of lookup error rate and lookup rate will be appropriate in at least two cases, 1) if the smart card is sufficient to support the lookup device, and/or, 2) a lookup-style disambiguation device will use At the receiving end of the communication activated by the smart card. For example, the user creates information with a smart card, and then transmits the information to another calculator through the telephone line, and after a period of time, uses a more powerful disambiguation device to perform disambiguation for the second time. Indeed, the second disambiguation does not need to be performed through the source producer, but can be performed through a second user, for example, the source producer's secretary.

However, this second code lookup error rate is still extremely low, for example, at this rate a very skilled typist will still hit the wrong key, the probability of this happening is about 1 in every 100 words . Wherever reasonable, both 16-letter key codes need to be accessible, so that a typist using the first 16-letter key code at 20 words per minute needs only Inquiries are answered every three minutes, compared to every five minutes using the second 16-letter keycode. When carrying out the same optimization procedure for the 9-letter key divergence code, we will find that akw bng cly dhx epv fim gr jot suz this group of codes only optimizes the search error rate. Here, the search error rate and query rate They are once every 116 characters and once every 4.4 characters respectively. This set of codes has been used in the standard design of the 9-letter keyed smart card shown in FIG. 16 . In order to optimize the search error rate and query rate with the guided random advance method, we can simultaneously establish a set of codes with a search error rate and a query rate of once every 109 characters and once every 6.2 characters, for example , am bnz cfi dhx evw gjr kosluy pqt this group of codes. It is worth noting that since lookups are not possible when using lookup-style disambiguation, touchability can only be discussed by reference to lookup error rates. Here we need to make an assessment of whether the lookup error rate is low enough to produce acceptable text. Despite the 9-letter key codes, the lookup error rate is comparable to the chance of a skilled typist pressing the wrong key, so in this context these codes can be considered touch-key codes. As a further example, since smart cards are most commonly used for sending SMS messages, making e-mails, communicating between pagers, etc., the level of text accuracy may be lower than that of transcribing the final text. From the above considerations, we get 9 to 16 keys to define the optimal range of keys for this application. If there are more than 16 buttons, it is not easy to retain the advantage of substantial button size while being installed on a smart card. Conversely, if the easy-to-find disambiguation device is compatible with the limited calculator functionality of the smart card, then a disambiguation code on a keypad with fewer than 9 letters may be a non-touchable keypad.

Feedback to Typists A smart card with an easy-to-find disambiguation device can be operated by a skilled touch typist over a telephone line with speech synthesis based on the symbols typed by the typist on the smart card, There is no need to get feedback on the communication progress from the card, and/or from the communication receiving device. However, it is also possible to provide feedback directly from the smart card, provided sufficient computer resources are available to provide feedback on smart card support.

It is noted here that fewer computer resources are required to provide efficient feedback than easy-to-find disambiguation device feedback. Although there is no disambiguation database related software on the smart card, and only the prototype electronic circuit board is used, the skilled user knows that a specific font can be directly transmitted to the display when the corresponding key is pressed, and the known The font is the letter most likely to be associated with the key. Take aw bi cx d ej fo g hv ky l mun pz qr s t as an example, the completed text is usually easy to read and understand for people. Taking the first line of Gettysburg's address as an example, using 1-block (single letter) data can be read as follows:

oour score and sehen kears ago our oathers irought oorth on this continent, a nea nation, conceihed in liiertk, and dedicated to the proposition that all uen are created erual.

An accuracy of this level is enough to provide a typist with a rough guideline about what he or she is typing on the smart card. This example illustrates that disambiguation can be achieved with an extremely small amount of memory; the only memory required here is to store the 16 fonts that will be displayed correspondingly when the 16 keys are activated. This processing method is scalable with the required computer resources. The more memory, such as 2-, 3-, the higher block probability can be stored and used as the basis for the well-known block-based disambiguation program, so it can display text modifications with increased accuracy to the user .

Although block-based disambiguation procedures are well known in the art, they have so far proven to be impractical. The following example will illustrate why: block-based disambiguation procedures are not efficient enough to disambiguate encodings that are too divergent. The block-based disambiguation procedure in this example is linked to a code that is sufficiently accessible and disambiguating enough to allow efficient block-based disambiguation. The traditional method has always been to use characters as the main disambiguation procedure rather than blocks as the main disambiguation procedure. However, applying the guiding principles of the present invention will make block-based disambiguation procedures operational and practical.

This example has further illustrated: 1) when using the guiding principles of the present invention, it is not necessary to use characters as the main disambiguation program, 2) when using the guiding principles of the present invention, no microprocessor is needed, 3) more One and possibly different disambiguation means can be used for the same communication system based on disambiguation codes. A character-based disambiguation procedure, or another disambiguation method, can be used on the receiving end of the communication sent by the smart card, and the smart card uses a simple block-based disambiguation method on the machine to provide feedback to the The user of the smart card.

As the length of the block used in the block-based disambiguation method increases, the correctness of text modification also increases. However, at certain block lengths the required memory capacity tends to approach that required by character-based disambiguation programs, and character-based disambiguation programs generally produce faster The reason for the better results is that you will usually choose to use a character-based disambiguation procedure, that is, if there is enough memory to support it.

Further Applications If more memory is available on the smart card than is required to store the disambiguation database and software, then the possible applications of the device can be greatly increased. For example, as long as there are a few more bits of memory, a user can query the information in the phone book through a suitable voice information system, and can query a phone number by typing a name and other similar attribute information on the smart card keyboard , and the telephone number searched can be stored in the user's memory so that it can be downloaded to another device with better performance later.

Query Minimization - Typing Devices Designed for Vehicles This application considers a situation where queries are the primary constraint in the design of a typing device. It is generally beneficial to reduce the query rate, which reduces the need to answer queries during the typing process. In some applications, however, reducing query rates is of paramount importance.

The query will be displayed on a video display for the most practical application of a typing device using a divergent code. When the user urgently needs vision, such as when the user is driving a vehicle, then based on safety considerations, the impact of vision on query evaluation should be minimized. Although inquiries are made via auditory devices, it is extremely important to minimize driver distraction. In addition, when driving a vehicle, it is usually necessary to hold the steering wheel with both hands at the same time, and it is best not to leave the steering wheel because of the need to operate a typing device. This object can be achieved by directly installing the typing device of this typing device on the steering wheel.

Referring to FIG. 17 , we will find that any input device can be mounted on the steering wheel 200 . Many steering wheels have pinches on the inner and outer layers to enhance finger grip. For such a steering wheel, it is natural to connect the first plurality of input devices 201 to the respective plurality of these pinches. When the driver grabs the steering wheel, the fingers of each hand will touch four of the first input devices 201 . The part where the driver touches the steering wheel may change from time to time, for example, when the driver turns the steering wheel at a large angle. Which 8 button groups the driver will touch at any moment can be confirmed by a position sensing device, such as a pressure sensing device and a combination of a simple electronic circuit board, which combination is used for skilled users will be obvious.

The second majority of input devices 202 can be mounted on the inner or outer layer of the steering wheel, and the thumb of each hand will touch one of the designated second input devices while the driver can hold the steering wheel. Which second input device the driver's thumb will touch at any one time can be determined via a suitable position sensing device.

With the above-mentioned keyboard installed on the steering wheel, it is a matter of course to select a code among the 8 buttons. The known buttons will be connected with the first input device touched by the fingers of both hands, and the two modes are switched. The button will be connected to the second input device touched by the driver's thumb.

Differential code selection Apply the guided random advance method advocated by this invention to the selection of the best code in essence, and only use the search rate optimization criterion, we can establish the following group of 8-character key buttons Encoding: aksz bcev dfi gmo hgt jnw luy prx. Here, the search error rate and query rate are once every 70.2 words and once every 4.1 words respectively. As noted in this entry, these ratios were calculated from our reference data using the simple character-major disambiguation procedure as the disambiguation method. To consider this encoding as touch-friendly its lookup rate may be overkill. A typist or driver typing 20 words per minute would be distracted from driving with a query about every 12 seconds, which may be too frequent to be compatible with safe driving regulations. On the other hand, a skilled typist may not be able to drive a car while typing 20 words per minute, so it is possible to bring the relationship between query and typing speed into an acceptable range for easy touch keys.

There are several additional strategies for reducing query rates that go beyond choosing a substantially optimal encoding, and these strategies can be used in combination. include

□ Increase the total number of buttons by increasing the number of buttons that can be activated with a single finger. It would be advantageous, for example, by adding an additional column of keys to the steering wheel, or equally setting each key to multiple positions, or using a coordinated approach, where more than two keys are synchronized Encodes a different subset of encoded symbols when pressed.

□ Eliminate queries when the likelihood difference between the low-likelihood decoder and the most-likely decoder is too large. The parameter controlling how close the probabilities should be between the low-likelihood decoder and the most-likely decoder must trigger a query, and the value of this parameter can be selected by the user. Such a mechanism would be of value in any application where the query rate is relevant to ergonomic standards.

□ Use a combined reconciliation/disambiguation encoding method, as detailed below. Use a more powerful disambiguation method than short-form character-based disambiguation.

Keypad Compatible with Existing Phone Keypads

In this application, the limitation on the number of keys is significant where the keypad needs to be compatible with existing telephone sets which typically have 12 keys. In this application, we need to reserve two keys for non-alphabetic symbols, such as space bar, cancel key, period key, and transmission termination key. Therefore, these 26 letters must be distributed on a maximum of 10 key buttons. In this application there will be a need for the lowest lookup error rate as well as the lowest lookup rate. We found that when using the best method and 10 keys separated equally as much as possible, we can find codes such as amq be cdu fiygqx hl jsv krz nw ot with a search error rate of 138 characters once and a query rate of 9.3 characters once. And it can optimize the search error rate and query rate at the same time. This is compared with the standard divergent encoding with a lookup error rate of 29 words once and a query rate of 2.2 words once, and an overall improvement of more than 4 times compared with the standard divergent encoding. In FIG. 18, the application of a 10-key code to a standard design of an existing telephone keypad is illustrated with an optimized lookup error rate and lookup rate.

We can also compare this 10-key coding with the 9-key coding proposed in the US CITE tegic patent and the CITE epo application EPO patent respectively. The first group of these codes, afg bkn jlo mqr dhi sux ptv cyz, has a lookup error rate of 86.5 words once and a query rate of 3.9 words once, the second group, rpq adf nbz olx ewv img cykthj su, has a search The error rate is 115 words once and the query rate is 5.2 words once. These codes are significantly inferior to the 10-key codes designed here for this task. When neither the US CITEtegic patent nor the CITE epo application EPO patent was born for the establishment of this divergent code, and these data did not exist when these codes were optimized (if they were indeed optimized), then we do not A conclusion can be made about the substantive optimality of these codes.

Another useful comparison is with 9-key encodings optimized in terms of lookup error rate and lookup rate. For example, we established the group am bnz cfi dhx gjr kos luy pqt with a search error rate of 109 characters and a query rate of 6.2 characters per encoding. Comparing these results, we found that the improvement obtained by the guidelines of this application has the following two sources: 1) using more than 9 buttons so that the search error rate and query rate can be improved, and 2) the search error rate and the search rate can be improved simultaneously, and Optimize query rate. The method described in this application will be extended to 11-key codes and 12-key codes, and we will find that the search error rate and query The rates are 215 characters per time and 10.1 characters per time, while the lookup error rate and query rate of a group of 12-key codes such as aw bn cky dhgq ef go ip jr lz mxsv tu are 313 characters per time and 13.2 characters per time.

That said, by sacrificing the use of the * and # keys when encoding non-alphabetic symbols, we can drastically improve lookup error rates and substantially improve lookup rates, thereby easily bringing standard phone-compatible keypads into ( Degree B) Easy-to-touch range. Whether these improvements compensate for the loss of the inability to use the * and # keys when encoding non-alphabetic symbols can only depend on the intended functionality of the device being built. It should be pointed out here that non-alphabetic characters can be coded in the coded sequence described above for smart card applications. Provided it is possible to use the * and # keys to encode non-alphabetic symbols, a particularly ergonomic design of using the * and # keys as transmission termination symbols follows in part as follows. Make # be a blank symbol=character key in the end symbol code, ## be=sentence key in the end symbol code, and ### be=transmit the end symbol code. The complexity of encoding such a symbol is inversely proportional to the probability of occurrence of the symbol. Depending on the application, the * symbol sequence can be used to encode other non-alphabetic symbols, such as the cancel key, @ (this is in the e-mail application), and/or as a mode conversion symbol.

The application of the telephone keypad arranged in English alphabetical order provides a solution to the severely limited keyboard design, which refers to a fixed number of keys, fixed position of keys, and fixed arrangement of symbols on the keys. As far as the keyboard is concerned. This issue arises in keypad designs such as 1) preserving as much as possible the final alphabetical order of standard divergent codes, 2) compatibility with existing standard telephone keypads, and 3) improved lookup compared to standard divergent codes Error rate and query rate. These constraints limit the degree of freedom in choosing a number of keys to be used as the basis for a differential code. For example, we can choose a dissimilar code in which a letter occupies 10 keys on the keypad of a telephone, then the * key and the # key can be used when encoding non-alphabetic symbols. Also, in cases where the standard divergent encoding uses as equal a separation as possible, we can choose an alternative separation that still obeys known constraints.

The individual permutations of these 26 units were separated into 10 groups associated with a unique ambiguous code, given the constraints of the alphabetic permutations. In the case of sufficient computer processing time, it is possible to evaluate the individual search error rate and query rate of these codes. A different and more efficient procedure would be to apply the optimization method advocated by this invention to this optimization-bound problem. The invention proposes that, in the absence of information suggesting the use of some of the more complex basic steps, it is necessary to define a most simplified basic step of possible encoding code groups. In the present case, the divergent code is a permutation list of 10 groups of letters, and all letters are contained in a group, and the letters will appear in alphabetical order. An example is ab cd efgh ij kl mn opqr stuv wxyz. Therefore, there will be 9 spaces in the group for separation. A basic step is to move a letter across a space. For example, if we choose the second space, in one basic move, we can get the abc d ef gh ij kl mn opqr stuv wxyz code by moving the letter c to the left, or by moving the letter b to the right a bcd ef gh ij kl mn opqr stuv wxyz encoding. When this particular code is known, all possible codes reachable by one elementary move from this particular code can be easily generated. Given this discovery and knowing the guided random progression described above, it will be apparent to a skilled user how to apply the optimization method advocated by this invention in the present situation. For example, applying this method, we will find out that the search error rate of ab c0d ef gh ijklm no pqr s tu vwjyz is 65 words once and the query rate is 5.8 words once encoded. This code is the ideal arrangement on the keypad of the telephone set shown in Figure 19. The error rate of this encoding should be compared with the standard divergent encoding whose lookup error rate is 29 words once and the query rate is 2.2 words once. After the comparison, it can be found that the search error rate has been improved by more than 2 times, and the search rate has been improved by nearly more than 3 times, and it does not affect the alphabetical order and does not affect the compatibility with existing telephone devices. . It is worth mentioning that the above discussion in terms of the choice of 11-key or 12-key alphanumeric encoding for a telephone set is applicable to this application as well; using separation optimization yields 11 and 12 Essentially the best code for key buttons.

This separation optimization method is obviously not only suitable for this application; for example, it can be applied to the smart card device discussed in the previous article, so as to obtain an alphabetical arrangement of 9-16 keys marked with alphabetic symbols The best encoding of the sequence.

Qwerty-like keyboard This method used in the previous application to make a keyboard that is 1) immediately compatible with standard keyboards and 2) immediately optimized for different ergonomic standards can be used to make a 1) standard-like Qwerty keyboard And 2) keyboards optimized for different ergonomic criteria. As was the case in the previous application we will maintain the design of the standard keyboard as much as possible by preserving the arrangement of symbols to key assignments and at the same time optimize the separation of these arranged symbols so that the lookup The error rate and query rate are minimized as much as possible. Such an application requires the further restriction that the letters will remain in the same row of keys as in the known Qwerty arrangement.

There is a Qwerty-like keyboard design sequence, that is to say, there are three columns dedicated to the letter keys, and a different number of columns, possibly one to ten columns. Obviously, with only one column, that is, three key buttons, the search error rate and query rate will be extremely high, and there is only one possible differential code corresponding to the arrangement of Qwerty keyboard symbols. This code is qwertyuiop asdfghjkl zxcvbnm, its lookup error rate is 2.8 words once and the query rate is 1.1 words once, a code of such poor quality is unlikely to be accepted as a human and practical application. As the number of columns increases, we will be able to find better and better divergent codes. Simultaneously, along with the increase of column number, the size that this device needs to arrange the key code, keeps the substantial full-size key code, also increases thereupon. Therefore, the Qwerty-like keyboard must be a compromise result between coding divergence and keyboard size. For example, if we wish to make a Qwerty-like keyboard the same size as a pocket calculator, but with full-size keycodes, we can use 7 columns, as shown in Figure 20. A substantially best easy-touch key code based on lookup error rate and query rate would be qwe r t yu I o p asd f g hjk l zxc vb n m, with a lookup error rate of 668 characters and a lookup rate of 35.5 characters at a time, obviously, this will belong to the easy-touch key type for many typists and keyboard applications of different degrees. In Fig. 20, this encoding is illustrated in an ideal arrangement. The keyboard-attached typing device shown in this figure is suitable for taking notes, writing e-mails, etc. It will be ready to type on and anyone familiar with a standard Qwerty keyboard will need no or minimal learning, and, despite being made with full size keycodes, will fit easily in a pocket.

According to the search error rate and query rate here, it is pointed out that the cost of attaching to the tradition is quite high, although it is only roughly attached to the tradition. If we can now have the auxiliary designation of letters to 17 keys, we will find that the code such as w r t bu gi ov p af s d ej ky 1 hz cx n mq has a lookup error rate of 7483 characters once, while the query The rate is 290 words once. This equates to one lookup error per 30 pages of typed documents and less than one lookup per page of typed documents. Such a device would be unimaginable if it were not of the touch-sensitive type.

Referring to Figure 21, we will see that the code can be designed with 18 letters at or very close to their Qwerty positions, which will be marked in bold. In this arrangement, in order to make the typing similarity between Qwerty keyboard typing and this optimized Qwerty keyboard typing the closest, the fingers should be placed on the original column, that is, the left index finger on (space bar) instead of Right index finger on (ej key). It can be seen that, by simultaneously placing the spacebar on the 'e' key on the original column, this design will take a big step forward in structural accuracy from the Qwerty design, and is more suitable for the most dexterous fingers than the Qwerty design. The degree of dependence will also increase relatively. Either of the differential codes can be optimally matched to a Qwerty (or other conventional) keyboard through an appropriate arrangement of key symbols.

The point to note here is that a very substantial advance in the utility of Qwerty similarity is made by allowing a slight difference from strict Qwerty arrangements. When the columns are only slightly spaced from each other, as in a standard Qwerty design, then many or most of the finger movements required to operate such a qwerty keyboard are the same or similar to operating a standard qwerty keyboard. The above illustrates the trade-off between ergonomic standards that preserve traditional order and ergonomic standards that preserve traditional functionality.

In view of the various standards that must be optimized and the different needs of users, in order to provide users with an optimized qwerty type keyboard and other keyboards that are optimized for ergonomic standards based on search error rate and query rate The choice between should be the driving force behind the practical application of this equipment. Being able to change button labels within the software would facilitate the above selection decision. In order to be able to do this, there must be a function on the key that can display more than one symbol at a time. The display function may be composed of a light emitting diode or a liquid crystal display.

Skilled users will find that the modern keyboard design method can be applied to other traditional keyboard design reservations and partial reservations, such as the azerty keyboard used in France.

The purpose of this type of application is to allow most calculator users to enjoy the advantages of different keyboards while keeping the existing hardware unchanged and at the lowest cost. These obviously include the advantages of one-handed typing and compatibility with different keyboards suitable for palms. Standard 101-key keyboards for workstations and personal computers usually include a set of numeric keys at the right end of the keyboard in a qwerty configuration. Usually there are arrow keys or functions that can be used to move the cursor near this set of number keys.

FIG. 22 represents the feasibility of synthesizing the divergent coding and moving cursor 601 functions optimized for common numeric keypad configuration 600 . The above-mentioned number key configuration 600 has 17 keys of different sizes in this example. Depending on other design constraints, all or some of these keys may be used for punctuation or other symbols. These design constraints may affect the selection of the number of keys assigned to letters, the distribution of letters and other symbols in different modes, and the like. Key features of this app include:

□Distinguished codes are assigned to multiple keys on the digital keypad

□Thumb-activated auxiliary typing function as the secondary replacement mode usage

Skilled users will find that the above-mentioned differential code assignment of the digital keypad can be achieved by relying on software; there is no need to use special function hardware. If the above-mentioned assigned keys are to include the properties of different codes, the key labels must be changed. Give the example of an entity, use the divergent coding under this setting, must select the divergent coding that the alphabet assigns to these 17 key buttons according to the existing standard corpus and the premise of finding errors and query rates. The code af bu cx d ej gi hz ky l mq n ov p r s t w shown in Fig. 22 has a lookup error rate of one lookup error per 7483 characters, and a query rate of one query per 290 words. This encoding has been discussed previously. The arrangement of the above codes shall be kept in the same alphabetical order as possible. This known encoding is not optimized for alphabetical order; it is only optimized for lookup error rate and query rate. According to the teachings of this invention, we can simultaneously optimize the search error rate, search rate, alphabetical order and/or other ergonomic criteria.

Figure 22 may be used to illustrate alternate mode usage in which the thumb activated assist typing function is secondary. According to this figure, we can assume that there are 4 key buttons to form auxiliary typing functions: up 602 , down 603 , left 604 , right 605 and other direction keys. These functions are usually based on 4 depressible buttons, but sometimes they are also based on key pads, joysticks, or any device that can generate multiple and different signals under the user's operation.

It is worth noting in Fig. 22 that several key buttons are marked with different symbols from the divergent codes, and in this example, these different symbols are numbers. These symbols can be obtained by depressing a specified one of the four auxiliary keying functions. Key-in functions on the modifier pad can be assigned as modes as follows:

□602 (Up) shift key for capital letters.

□ 603 (Down) Number/punctuation mode.

□604 (Left) button is marked with a left symbol.

□605 (Right) button is marked with a right symbol.

It's worth mentioning that 1) we can use other types of modifier keypad symbol and/or pattern assignments based on the teachings of this invention, and 2) more symbols and patterns can be assigned to more complex secondary keystrokes on the function module. We'll discuss schema assignment for symbols in more depth in another application. This discussion will apply to this and other applications.

OBJECTS AND ADVANTAGES OF THE 13-LETTER KEY-TYPE CODING Many related applications of the teachings of this invention benefit from the surprising advantage of optimization of divergent coding; that is, half the number of keys are dedicated to extremely relevant symbols. If we choose [a-z] in the English alphabet as extremely relevant symbols, then the ideal number of keys is 13. The 13-letter touch-tone code in English includes the following amazing advantages:

□ Easy to touch keys,

□Ergonomic, touchable key typing, non-differential text typing method,

□Ergonomic, touch-key typing, inquiry method,

□Compatibility with standard keyboard configurations (qwerty keyboard, numeric keypad, telephone keypad),

□Provide retention from one-handed to two-handed typing skills,

□Provide a mouse/keyboard combination,

□Provide devices to reduce typing damage.

More purposes and advantages can be seen from the following detailed specifications.

Ease of access From Figures 11 and 12, it can be seen that in the character-based ambiguity code elimination method, even skilled typists believe that the ambiguity codes on the 13 key buttons have a high degree of ease of touch. Using the aforementioned guided random advance, we will find that the search error rate of the code aw bn ck ef go hvi p js ly mx qt rz appears once every 515 words, and the query rate appears once every 21 words. It is the easy-to-touch key degree of C grade. Figure 25 shows the ideal arrangement for this encoding. As mentioned earlier, changing the variables that control the importance of a query and whether it gets the user's attention can reduce the query rate even further. To the extent that queries are not used at all, accessibility is governed by the lookup error rate. For the above codes, on average there will be approximately one lookup error per two pages of typed text. This rate is far less than the error rate of a skilled typist. So the 13-key button type coding is applicable to various touch-key typing tasks and users.

Ergonomic, non-disambiguating text entry for touch typing It would be very convenient to be able to use any disambiguation keyboard and any disambiguation code disambiguation method, such as numeric entry in a disambiguation database, in a non-disambiguation information entry method. In all uses of differential keyboards, the more ergonomic the non-differential text typing method is, the better it is, that is to say, the easier it is to operate, the better. The divergent keyboard is originally used for typing with tactile keys, and it would be more convenient if the keyboard can be operated with tactile key typing and non-discriminative information typing.

In order to achieve the non-discriminative text input method using few key buttons, a common strategy is to adopt the coordination method. In order to achieve the requirement of using the least number of keys, coordination method designers are constantly studying the form that can simplify the coordination method with a sufficient number of keys. According to the basic comprehensive proof, the complexity of the coordination method cannot be more than 2, that is, there is no need to activate more than 2 input functions at the same time to achieve non-discriminative symbol encoding, and the number of keys cannot be less than half of the number of encoded symbols . The present invention contrasts with the above technique and points out that the number of keys cannot be less than half of the number of coded symbols if the simple function of unambiguous symbol coding exists. Taking a more unique example, the current invention instruction requires at least 13 keys to represent letters [a-z], and at least one mode switching key. When this mode conversion key is used simultaneously with other keys, the letters related to the above keys can be encoded uniquely and non-discriminatively.

On a divergent keyboard, some keys can represent many symbols with a single press. If the same keyboard is to be used in non-differentiated mode, the above-mentioned single key must be combined with at least one other key, perhaps using this key, to single-select each symbol associated with this key. In terms of ergonomics, the simpler the combination of the above keys, the better. In terms of accessibility, the same key combination would be ideal for unambiguous typing of all symbols. To reach the above two criteria, 1) the number of symbols assigned to each different key must be the same, and 2) the number of symbols on each different key must be reduced. In summary, the number of these ideal standard keys for differential coding should be half of the total number of symbols on the differential keys. For example, the above standard means that 13 is the ideal number of keys representing 26 English alphabet symbols.

The above findings can be used as a reference according to Figure 23 and the 13-key button type divergence representation of English letters, and use the non-disparity text typing mode that conforms to ergonomics and tactile key typing on the divergent keyboard that conforms to ergonomics and tactile key typing discussion. As can be seen in the above figure, each distinct subset of representative letters 700 that represent letters 700 encodes only two letters. The touch-typing keyboard also includes a mode key 701 and the function of converting divergent and non-discriminative typing methods. The function of switching between divergent and non-differential typing methods can be controlled by software, and the mode can be switched according to the situation at that time, or a certain key can be assigned according to the conversion mode, or a special type of typing function, such as the mode switching using double keys key button 701 . In non-discriminative typing mode, if the key belonging to 700 is activated almost simultaneously with the key 701, activation of the key 701 will encode one of the two symbols associated with the key 700.

It is ideal to describe the symbol of the 700 key button as a left symbol and a right symbol, and the left symbol is marked on the left side of the key button and the right symbol is marked on the right side of the key button. Without loss of generality, the left symbol to the left of the key is associated with the activation of the 701 key in order to achieve non-ambiguous text entry. When any key button in the 700 key button group and the 701 key button are activated simultaneously, then the above-mentioned left symbol will be subject to non-differential selection. If the same button in the above-mentioned 700 button group is not simultaneously activated with the 701 button, the right symbol will be subject to non-discriminatory selection. The above-described non-differential text input method is also applicable to keyboards that can incorporate other modes.

Adapting to touch-key typed query methods and even optimized ambiguity coding, unlimited computer processing power plus undeveloped artificial intelligence disambiguation technology, ambiguity sequences will still be generated when text is typed, and need to use The intervention of the parties to achieve the best effect of eliminating differences.

A true touch typist can keep his eyes on the typed text or the original text instead of looking at the keyboard. An ideal arrangement for a touch typist would be to arrange all interactive interpretation of divergent sequence queries so that 1) they do not interfere with the typist's attention to the screen, and 2) they are accessible from the keyboard in a simple and fixed manner. Reply to all inquiries. To achieve the above purpose, a series of candidate characters can be accessed by a reserved key and different characters can be selected on the screen. The user can operate the scroll bar buttons to scan the candidate characters word by word, and when other than the scroll bar buttons are pressed, the characters in the scanning column are in the state of being selected.

With reference to Figures 23 and 24, we will discuss in more detail the software that forms the basis of the touch-key typed search method and the visual display controlled by this software. In a first step 800 a query is detected, that is to say the disambiguation function finds more than one set of meaningful decoding sequences in the database corresponding to the entered coding sequence. The function after entering the query mode will divert the user's attention to the displayed query code. Such functions may be visual functions like box 702 in FIG. 23 . Possible decodings are then listed in order of their likelihood (step 802). Afterwards (step 804) the highest likelihood decoding is displayed on the screen in the attention function area as described above. The software is then in a ready state, waiting to receive the input of the scroll bar key or other keys (step 806). If it is the typing of other keys, the function of diverting the user's attention will be removed (step 808), and this code will be added to the previously typed text (step 810) and then get back to the different text typing method. Conversely, if at step 806 a key input of a scroll bar key is detected, then the database is used 812 as a reference to test whether meaningful decoding can be obtained. If so, the existing decoding is replaced by the next more likely one (step 814), and then returns to step 806. If there is no more likely decoding, the input device enters the above-mentioned non-ambiguous text input mode (step 816), and the decoding sequence is entered in a non-ambiguous manner, and the function of diverting the user's attention will be removed (step 808), and this encoding will be added to the previously typed text (step 810) and then get back to the different text typing method (step 818).

Although the above methods representing other options are described in terms of adapting to the TTY type query method, the same method can be applied to other situations as well. For example, when "codes" represent characters with associated meanings, the database is a thesaurus, or when "codes" represent the various meanings of a character translated into a foreign language, the possibility of codes is provided by automatic translation software provided.

Cross-platform design preserved to meet ergonomic standards: mouse/keyboard Assuming that a user with a handheld device such as a personal digital assistant with a built-in one-handed typing keyboard, at the end of the typical day, it is likely that he or she will also use hands with keyboard to calculator. If the user is to effectively use touch typing on both devices, the movement patterns for the one-handed and two-handed keyboards must be as similar as possible. The shorter the time spent switching between using two keyboards, the more important it is to preserve typing skills. This invention provides a device for fast switching between one-handed and two-handed keyboards and can achieve the retention of cross-platform typing technology.

The invention relates to a one-handed keyboard suitable for inputting tabular data in a spreadsheet or a network. This invention can be usefully interacted with programs, such as video games or graphics programs, 1) that require fast cursor-to-type replacement, and/or 2) when the appropriateness of a symbol's typing depends on the position of the cursor on the screen. When using the qwerty keyboard and mouse that come with a standard calculator, the user must move his hand away from the keyboard to operate the mouse. In tasks involving rapid and continuous typing and mouse operations, such as marking a design on a calculator screen or filling in a form such as a web page, the interaction between the mouse and keyboard can be very slow and slow. trouble. In this invention, a one-handed keyboard is mounted on a structure that can slide on a flat surface, so that it can have the functions of a mouse and a keyboard. Although users may think that two-handed keyboards are better for typing tasks, a more similar key arrangement between the two keyboards would be ideal for mixed use between one-handed and two-handed keyboards. Typing skills transfer seamlessly.

From Fig. 25, 28 and 26 we can find out, select and can be arranged on the divergent coding on one-handed keyboard and make the action of finger and thumb (the situation of right hand among the figure) no matter be on one-handed or two-handed keyboard. The same, and the typing action of the other hand is also similar to that of the selected one-handed typing.

This design strategy is as follows:

□Choose 13-button button code with sufficiently low search error rate and query rate.

□Select the appearance of the arrangement for the above 13 buttons. An ideal arrangement would be 5 buttons on the top row, 5 buttons on the middle row (base row), and 3 buttons on the bottom row.

□Select which hand to use to activate the one-handed keyboard.

□When arranging the keys, it needs to be determined according to the selection of the left and right hands in the previous step. and:

——Use the proportion of the reference line as the maximum value.

- The specific gravity of the most commonly used finger is the maximum value.

——The proportion of the most upward is higher than the proportion of the most downward.

Then, pair all the keys activated by the left hand with the keys activated by the right hand to obtain the arrangement of the two-handed keyboard, and the keys of the left and right pairs must use a straight line drawn from the center of the keyboard from bottom to top as the plane of symmetry.

·Select one of the two symbols assigned to the original 13 buttons, and connect the selected symbol with one of the paired buttons in the above steps. This step can be carried out in favor of a one-handed keyboard or an associated two-handed keyboard.

- Favor one-handed keyboards: Arrange high-used keys on the chosen side of the two-handed keyboard, and less-used keys on the other side of the keyboard.

——Being partial to two-handed keyboards: Arrange the keys with high or low usage on the same side of the original 13 keys on the two-handed keyboard, so keep them on the same side

- The total number of possibilities of using both sides of the keyboard is as equal as possible.

Starting with the choice of 13-key coding for this application and a preference for right-handedness to illustrate its principles, the keyboard design shown in Figure 25 can be established. When this one-handed keyboard is used as a two-handed keyboard, the resulting design is shown in FIG. 26 . On this keyboard, about 84 percent of the letters can be typed with the right hand, and about 16 percent with the left hand. This asymmetry is ideal in a situation where most keystrokes will be performed in almost the same way, whether using a one-handed or two-handed keyboard.

On the contrary, if the keyboard is typed with both hands, and only one-handed keyboard is used occasionally, it can be said that the proportion of reliance on both hands is as equal as possible when operating the two-handed keyboard. This goal can be achieved by an alternative design of a two-hand keyboard shown in FIG. 27 . It is worth mentioning that whether you use your right hand or your left hand to type on this one-handed keyboard, 50% of the typing action on the two-handed keyboard is essentially the same as that on the one-handed keyboard.

It is also worth mentioning that there are ²¹³ different ways to pair the corresponding left-hand and right-hand keys of a two-hand keyboard in terms of 13-key differential coding. This number is small enough to evaluate each pairing with the corresponding weight of the hands, and the appropriate setting to choose from will depend on whether the desired weight is the most symmetrical or the most asymmetrical or some Depends on the middle value.

Referring now to FIG. 28 , a detailed illustration of a one-handed keyboard, we can see how these objects are achieved by placing several keys 300 on the keyboard, a thumb-operated typing device 301, a mouse 302, and a palm grip. Use 303 and a display 304 to achieve. The keyboard can be further equipped with a communication device to facilitate the symbol selection process transmitted between the calculator and the keyboard. This can be a wired or wireless communication device, such as an infrared communication device. The keyboard can be safely moved on a supporting device, such as a table, so that the keyboard can be operated by sliding on the device with pressure from the palm of the hand. Placing a palm grip on the keyboard would be ideal for using palm pressure to effectively move the keyboard. In the case of a mode in which the keyboard can be effectively moved in any direction with only slight pressure, the device indicated to operate the keyboard with palm force would be an ideal device. For example, the pattern could be a groove on the keyboard where the palm rests securely. By moving the keyboard in this way, the fingers operating the keyboard can be used to effectively operate the keys, even when the keyboard is moving. It can be seen that this keyboard can be applied in the situation of playing computer games, that is, when it is necessary to input actions and symbol sequences at the same time.

It is worth noting here that when the device is performing the function of a mouse, its appearance is very different from that of a mouse. Its mode element is determined according to whether the hand structure of touch-key typing is carried out on a comfortable position. Therefore the device must be much larger than a standard mouse, and the means for moving the device will be substantially different.

To transmit keyboard actions 305 to the calculator, the keyboard in question has been fitted with a movement sensing device, such as a trackball, which is well known to skilled users. Ideally the keyboard 305 could be further fitted with a biasing device, such as a spring, to lift the keyboard from the support when hand pressure on the keyboard is reduced, thus making it easier to move. Conversely, when substantially full hand pressure is applied to the keyboard, the keyboard can remain relatively stable mounted on the support device while the typing action is assisted.

In this way, the two-handed keyboard can be used for typing without being affected by the mouse movement for a long time, and the one-handed keyboard can be used for fast and variable typing of the mouse movement.

The visual representation of the keyboard is noteworthy when using a single device that supports more than one divergent code or more than one mode, at any given time having a representative representation of the current associations with keys and symbols appear on the display It is quite practical for users. Such a typical representation would be beneficial to any typing device, especially typed devices, because some or all of the key buttons of such devices would be visually inaccessible in use (especially during operation). refers to the key next to your finger). Therefore, any display combined with key buttons will be of limited functionality to the touch typist. The most useful visual representation is when the physical design of the keyboard is implemented on a visual display. Such a device 304 is shown in Figure 28 and can be combined with many of the applications described herein.

Reduce Typing Injuries Many keyboard users suffer from typing injuries (repetitive stress syndrome). Countless keyboards have been designed to reduce the stress of repetitive motion during typing. It has long been recognized that the most effective way to reduce typing injuries is for typists to take regular breaks during the typing process. However, this is not practical because typists are often pressed for time when it comes to completing their typing tasks. The one-handed keyboard just described provides a solution for this. Although the current one-handed keyboard is operated by the right hand, the same design method can obviously be used for a left-handed one-handed keyboard. Each of these keyboards encodes all the same symbols. A therapeutic typing device equipped with a right-handed keyboard and a left-handed keyboard, such as a left/right-handed mouse/keyboard described above, can be operated with either the left hand or the right hand. With such a pair of keyboards, a user who wishes to reduce repetitive stress injuries can use one of the keyboards to type for a period of time, such as 15 minutes, and then switch to the other keyboard for the next period of time, and so on. The user allows each hand to have a rest period and will not reduce the amount of typing. This typing device can be equipped with a locking device to lock one of the keyboards so that it can be used alternatively. It is worth mentioning that after the treatment is completed, the user can return to the two-handed keyboard mode without having to relearn the required skills.

Foldable personal digital assistant (PDA) We can see that in a smart card application described above, the text screen of the differential code typing device is placed on the keyboard part operated by the fingers of one hand and both hands, and the thumb can be used to operate Alternatively the location of the keying device, ie on top of the thumb or between the thumb and finger-operated keyboard area, would be convenient and ergonomic. The current application is to consider a typing device using the same concept in a larger scope while combining a folding concept to design a twice-folding information device. Such a device will be foldable when unfolded. Ergonomically performs different functions when folded once and twice.

This double-fold design is an astonishing result of a divergent code. Notably, a typing device constructed in accordance with the methods advocated by this invention enables a keyboard that is simultaneously (1) capable of efficiently encoding natural language, (2) employing substantially full-sized keys, and (3) The size is suitable for carrying in a pocket or a small tote. This application is based on the construction of a palm calculator device using substantially the same basic components, namely a keyboard size designed with a different code, and the components referred to are in several according to the needs of the user at the time Will vary depending on the case is adjustable. These primary assemblies can be connected to each other in folded and/or detachable form in each of the different states. In this way, the calculator device can play the role of a notebook calculator, a personal digital assistant PDA, a telephone, a game machine and so on.

Referring first to FIG. 29, we specify that a twice-folding calculator is composed of four substantially identical parts, each of which is used to perform a specified function and is folded and/or foldable from each other. Remove the connected form. Figure 29 illustrates such a device in folded mode. It can be seen that the first side of one of the components 900 is used as the first visual display, the first side of the component 901 is used as the first keyboard, and the first side of the component 902 is used as the second for. Ideally, the keypad design on the first and second keypads would be a 13-letter keypad, although there are many other possible options. The first surface of the last component part 903 is used as a pair of thumb switches for mode switching, so as to be operated in combination with the first and second keyboards here. The first keyboard here is designed to be operated with the right hand. It will be apparent to the skilled user that a similar pattern exists for typing with the left hand, and this similar pattern can be achieved by simple rearrangement and reinstallation of the four components. Indeed, a two-handed keyboard can be achieved by rearranging these four components, as shown in FIG. 33 .

Shown in Figure 30 is the bottom of an unfolded double-fold calculator. Components 904 are a telephone keypad 905 and a corresponding second visual display 906 . Component 907 is a third visual display and component 908 is a third keypad. Elements 904, 905, 906, 907 are the second sides of elements 900, 901, 902, 903, respectively.

Folding the calculator along the line 908 shown in Figs. 29 and 30, we get the pattern shown in Fig. 31. In this mode, the third keyboard is used for typing, and the third visual display is used for displaying corresponding images. The keyboard here is designed as a 12-key keyboard, but many other options are possible. This mode can also be used when the user is unable or unwilling to fully unfold the calculator due to time and space constraints. It can also be used to support a different functionality than the fully deployed calculator, such as for gaming.

Finally, the computer can be folded according to the fold line 909 shown in FIG. 31 to form the double fold pattern shown in FIG. 32 . This is a mode especially suitable for portable calculators; in this mode the unit is pocket sized. In addition, the function of the telephone is also included in this double-folding mode. For many users, this will be the mode in which the device will be used most. It is to be noted here that we have illustrated a telephone which uses the divergent coding in the previous application, however many other options are possible.

Statement again: If it is not due to differential coding, it is impossible for us to design a portable communication and computing device that can be switched between telephone, personal digital assistant and notebook calculator.

Skilled users will find that the different modes and uses of the device will be further expanded if every primary component, including the keyboard, is touch-sensitive from the screen. However, the tactile feedback of standard keyboards and depressible buttons is lost. Many other options are possible while consistent with the principles of this invention.

Typing Device Software Application Including a Touch Screen This invention enables both software and hardware applications. In particular, the method of the present invention can be used to design typing devices for devices that include a touch screen, such as the Personal Digital Assistant series manufactured by 3Com Corporation and distributed under the PALM PILOT trademark and other trademarked products. For purposes of illustration, we will focus on the PALM PILOT family of products, which include handheld calculators that can or already implement one of these applications, however the approach presented here is applicable to any of the typing device.

Referring to Fig. 34, we noticed that the PALM PILOT series devices are generally composed of the following components: a touch screen 1000, a tactile sensing area 1001 for entering fonts through handwriting recognition software. The tactile sensing area referred to here may be a sub-area of the touch screen, or may be applied separately.

One of the main and amazing features of this application is that when using a touch keyboard on a device consisting of a touch screen, we have designed a new user interface for the information device, whereby the keyboard does not Requires a tug-of-war with the app over limited screen real estate. The same touch screen can also be used for applications as well as this keyboard.

An important finding is that, if the keyboard is low-touch, there is no need for the user to actually display the keyboard. The user's finger "knows" where the keys are without the need for visual indication. In this way, the keyboard can be used to type data while every application is displayed on the touch screen. Furthermore, if the keyboard is touch-sensitive, it can be used to type high-quality text even when there is no screen real estate to provide user feedback on queries.

Referring to Figures 34 and 35, some of the unique features of the PALM PILOT series of devices that have been described in this application are:

□The touch screen 1000 can easily replace the functions of other keyboard designs.

□The touch screen 1000 can display images with different brightness and colors.

□The location of the typing area 1001 is a certain distance from the touch screen, or a non-central area of the touch screen.

□Use this personal digital assistant to execute different programs, such as sequencer programs or communication rate programs, and these programs will occupy the same touch screen space as the keyboard.

Touch screens can easily replace the functionality of other keyboard designs in this application to implement a given keying device to represent many different symbols or groups of symbols, depending on the keyboard mode at any given time. When the touch screen is used on a keyboard, each input device is connected to a designated area of the touch screen. The dual function of the touch screen as both a visual display and multiple mechanical keying devices is used to assign different functions and different labels to each keying device according to the mode. However, it is worth mentioning that the same effect can be achieved by using traditional depressible mechanical key buttons with their individual display devices installed on each mechanical key button. As such, the method of mode switching indicated here with reference to a device consisting of a touch screen can be applied to devices consisting of mechanical buttons, such as many other devices described herein .

Mode Selection One codable strategy for increasing the number of symbols in a keyboard design given that the number of keys is known to be fixed is to increase the number of keys as the mode switches. Pressing a mode shift key changes the symbol encoded by a number of other keys. A standard example is the shift key on a typical typewriter keyboard that encodes symbols by changing letter keys from lowercase to uppercase. Uppercase letters could in principle be coded with a different set of keys than those for lowercase letters, and it would be advocable if, in normal communication, uppercase letters appeared with the same frequency as lowercase letters. s Choice. Simultaneously, in principle, the uppercase and lowercase letters that can be reached in different modes do not mean that an uppercase letter must be connected with the same button of the corresponding lowercase letter. In practical application, the same key assignment is also chosen, because it is very traditional, easy to understand, and there is a statistical relationship between upper and lower case letters.

It follows that the following three principles govern the designation of symbols to modes and keys within modes: degree of association with statistical relations, degree of association with traditional relations, and conceptual relations between symbols. The problem of pattern design is most serious when using differential coding in the design of typing devices, because several keys must already have the function of encoding more than one alphabetic symbol on each key, and enough The number of keys used to encode symbols is usually very limited. However, such a method, which has been applied to create divergent codes for alphabetic symbols, can equally be applied to non-alphabetic symbols such as punctuation marks in cases where these non-alphabetic symbols are closely connected.

The symbols that need to be encoded by the keyboard will be divided into several subsets corresponding to the mode. Depending on how much user manipulation is required to achieve the individual modes and/or how often symbols are used within each mode, the modes will be at least partially permuted. Then we can call them first-class, second-class, third-class modes, depending on the increase in the number of operations required to achieve the individual modes and/or the decrease in the frequency of symbols used in the mode.

It is ideal to place alphabetic symbols in single or several first modes. Subtle design issues must be related to the way the non-alphabetic symbols designate the patterns and the spatial design arrangement of the patterns.

The first statistical criterion to consider is the frequency of non-alphabetic symbols. Some non-alphabetic symbols, such as punctuation marks and numbers, are important to communications that may have equal or greater usage than alphabetic symbols. These punctuation marks are candidates for inclusion in any valid keyboard design in first or second class symbol sets. The next statistical criterion to be considered is connectivity due to contact of non-alphabetic symbols with other non-alphabetic symbols. There are traditional and conceptual relationships between certain non-alphabetical symbols and other non-alphabetic symbols, for example, the symbol (left parenthesis) is connected to the symbol (right parenthesis) because the two symbols are in Together it makes sense. This symbol is connected with this symbol because the two symbols have similar meanings, word ending or sentence ending. These are examples of global usage that are quite familiar to most speakers of the language, including many languages such as English in these cases. Also can consider other relatively native relations on the keyboard design of special purpose, for example, the relation between: and/:/ is generally used in the webpage address (URL or website address) of Internet.

Non-alphabetic symbols may also have statistical, traditional, and conceptual relationships to alphabetic symbols. For some symbols, it is possible to analyze their statistical relationship to each other through a reference corpus. For other symbols, since they almost never appear in text, analyzing the statistical relationship of these symbols requires user studies or specialized software. Available "backspace" (removal key), "page up" (previous page key) and other symbols for editing view or otherwise processing text are used as examples.

Where reference data are derived in this way, the next symbol-to-schema specification step is to arrange the symbols in a way that best fits statistical, traditional, and conceptual relationships. A further limitation that can also be considered is the memory potential of the arrangement. Ideally, all symbols are arranged in a "meaningful" way in all modes, that is, the pattern of symbols is simple, familiar, and preferably visually structured. Even highly trained touch typists may fall back to using visual scanning patterns to find symbols on the keyboard that are not used often. In this case, memory potential may become a major consideration in permutations dedicated to lesser-used symbols. It is worth mentioning that mnemonic potential can be quantified by the experimental protocol set up for the secret memory task familiar to psychologists.

To illustrate this method, there must be a standard design for alphabetic symbols [a-z], numbers, and 32 non-alphabetic symbols

~`! @#$%^&*()_-+＝{[]|\:;”<,>.?/

This is found on a standard keyboard as specified. As shown in Fig. 36A-C, such an arrangement is made up of three mode switch buttons, three modes each including 16 symbol keys. This configuration is designed for the PALM PILOT series of devices. This design has not been proven optimal by psychological testing.

The first mode here, (Fig. 36A), includes a divergent code for a letter, a space (space bar) / backspace (cancel key) button, a basic punctuation mark, and a forward or backward shift mode button.

The second mode here, (Fig. 36B), contains several additional punctuation marks, and the arrangement of these symbols will be 1) a shift (transfer) key button is connected to the symbol with related meaning, for example, left and right brackets, Or, linking to a symbol shape if there is no relevant meaning, which will help remember the symbol's location. Here a tradition applies, namely that the hard symbols, the more water chestnut ones, are on the left, and the soft symbols, the more curved ones, are on the right. Since all or most of the keys have exactly two symbols, the ergonomic disambiguation device as described above can be operated in each mode.

Selectable Transparency of the Keyboard At present, the general application of the keyboard displayed on a PALM PILOT series touch screen device is that the keyboard occupies a part of the screen, and the remaining screen area is dedicated to an application that receives the keyboard input , such as an address book application. Given the extremely limited display area of the device, both the keyboard and the application must be very small as a result of the co-existence between the keyboard and the application. The keyboard used in this device is not suitable for tactile typing, nor will it be tactile due to its size. However, a keyboard produced using the inventive method will be sized as a typing device, albeit in a limited location on the touch screen of a PDA. An important discovery here is that, where the keyboard is touch-sensitive, then the keyboard will not need to be displayed to the user. The user's fingers "know" where the keys are, even if they can't actually see them. Therefore, the keyboard can be transparent and occupy the entire touch screen, while the application program can be opaque and occupy the entire touch screen at the same time. With such a display method, the user can directly type on the application program. In Fig. 35, there is a keyboard and an application program 1002 displayed in this way 1003, and what is shown here is that a drawing program is performing a drawing function. It is not possible to draw a transparent keyboard, the diagram shows the keyboard in gray and the application in black. Indeed, it is possible to let the user choose keyboard transparency, for example, according to her or his touch-typing skills.

It has been pointed out that different modes may contain symbols of different familiarity to the user. To account for these differences, keyboard transparency can also be adjusted by mode function, decreasing transparency as symbols in the mode become less familiar. It is worth noting that the same effect of distinguishing the keyboard from the application can be achieved by adjusting other visual factors, for example, adjusting the color of the image in addition to adjusting the transparency.

Synthetic Harmonized/Different Keyboards An important finding against this view of the invention is that a harmonized pattern requires only two key buttons for substantially simultaneous operation, such as a harmonized pattern that makes a Qwerty-style keyboard encode capital letters, It will be immediately learnable and available to a large user base. However, traditional methods have proven that coordination styles more complex than this will not be widely accepted.

It is worth noting that there are two main traditional methods to create a typing device with a small number of keys: the harmonized method and the divergent coding method. One of the viewpoints of this invention is to illustrate how to combine these two methods to bring out the best in each other.

We can distinguish that there are two coordination methods 1) reserve a key or several keys as a method of forming a harmonic function, the familiar method of typing capital letters through a coordinated combination of a shift key and a letter key Shifting key buttons is an example, and what we usually refer to as this type of key buttons refers to shifting key buttons. 2) A coordination method formed by substantially synchronously operating several letter key buttons. The first of these methods is used in this application, and the second of these methods is used in the next application.

The main insight of this aspect of the invention is that a group of keying tools operated substantially simultaneously by the user can be instantly combined into a single command. Therefore, a group of button actions will not be easier or more difficult to control than a single button action. However, a group of button substantially covers more information than a single button, that is to say, it can be used to create easy-to-operate and A low-discrepancy code and a typing device built according to the code. It follows that, to make the keyboard simple to operate, coordination must be performed on no more than one group of keys in substantially simultaneous operation. In this application one of the groups will be reserved for forming coordination keys, while the other group will be a key corresponding to at least one decoded symbol. While less frequently used symbols may still require substantially simultaneous operation of more than two key buttons, frequently used symbols may be associated with a single keying tool without exceeding the scope of this invention.

It would be ideal if at least one of the decoded symbols is an extremely correlated symbol. Thus, if the formation of a coordination is incomplete, ie when the shift key button is pressed when it should not be or is not when it should be, the disambiguation software can be used to correct this error.

It can be seen that according to the principle of this application, a priority can be composed in any way when performing coordination and divergence coding procedures, and the ideal combination is as follows

□ No more than two keying tools are required to operate substantially simultaneously to encode any substantially possible symbol.

□ Lookup error rate and/or lookup rate are optimized.

□ Coordination can be done via a mode shift key.

□(Ideally) The probability of using the mode shift key is the lowest.

Astonishing and synergistic results can be achieved when the combination of harmonized and divergent coding procedures is followed in accordance with this principle, which will be achieved in this application by an integrated harmonized/discriminative coding procedure applied to telephones containing standard divergent codes illustrate. We can already see that the lookup error rate and lookup rate of the standard divergent encoding are quite poor. Therefore, it is quite unusual to use a comprehensive tool to make a keypad using standard divergent codes easy to touch. The goal of this application is to create a keyboard as described below

□Easy-touch keys, fully compatible with standard telephones using standard divergent codes,

□Easy to operate,

□Easy to learn,

□And it uses the least key button action.

Shown among Fig. 38 is exactly a standard telephone set using the standard divergent encoding standard. As can be seen in this figure, several key buttons 10000 are used for encoding letters and numbers, and here there are eight in total. The two key buttons 10001, 10002 only encode numbers, while the two key buttons 10003, 10004 respectively encode the non-alphabetic symbols * and #. In this application, one of the keys selected from the group consisting of 10001, 10002, 10003, 10004 will be used as a mode shift key, ideally 10001 is coded for the number 1. The selected key will be called the shift key, for reasons that will become apparent. When the key button 10001 is held by the left hand, it can still be operated conveniently with the thumb of the left hand, while the right hand is used to operate other key buttons. For an application where the right hand holds the phone and the right thumb operates the shift key, key 10004 can be used as a shift key.

Each of the 10,000 keys has a corresponding letter, and these letters will be divided into two subsets, which we call the shifted group and the non-shifted group respectively. The designation of letters and key groups (transition group, non-transition group) will be as follows

□Lowest lookup error rate,

□ The lowest query rate,

□The transfer group, without losing generality, will have the characteristics of a letter matching a button,

□Lowest transfer button operation probability

In general, optimizing synchronization in terms of lookup error rate, lookup rate, and other ergonomic criteria means that there will be some compromises. For example, by eliminating a single letter per key button, so that the number of letters in the transition group can vary from key to button It is possible to achieve better lookup error rate and lookup rate within the limit of . However, the regularity of separation between shifted and non-shifted groups makes the keyboard easier to learn and conforms to ergonomic standards, which is a priority here. The degree of learning can be further improved by selecting individuals that are also easy to remember when several sets or by memory. However, this choice may affect the lookup error rate as well as the lookup rate.

There are a total of 11664 different transition/non-transition pairs that are restricted to contain only one letter per key in the transition group. This number is small enough for its ergonomic properties to test all possible combinations.

Figure 39 presents the results of testing all the above-mentioned 11664 codes and standard divergent codes (SAC) in the form of a dot plot of the search error rate versus the query rate. It is worth mentioning that while all encodings are better than standard divergent encodings, most are only slightly superior. However, this is a rather large distribution. Take the best code CEHLNSTY as an example. Its lookup error rate is one lookup error per 431 characters, and the lookup rate is one query per 21 characters. This is in terms of lookup errors 15 times better than standard divergent encoding, and 10 times better than standard divergent encoding in terms of query. The best code is AbC dEf gHi jkL mNo pqrS Tuv wxYz, and the parts that use uppercase letters are the elements that move into the code group. It is worth emphasizing again that this encoding is the best encoding we have available for the reference data. While this encoding is still the best choice for data from many other English corpora, other sources may yield different optimal encodings. The comprehensive coordination/disparity coding method can be applied to any divergence code, and this divergence code is not limited to whether it is a standard divergence code or English letter arrangement, and there is no 8 key buttons or equal separation method as much as possible restraint. If there is more freedom to choose the code, the quality of the integrated harmonized/differentiated coding method can be greatly improved in terms of search error rate and query rate, and other ergonomic criteria such as minimizing the use of shifting keys can be achieved. Advantageously combined with optimization of error rate and query rate. Such an optimization step consists in pursuing full compatibility with existing telephone systems. Although beyond the scope of the above application, it is still covered by the present invention.

In order to promote the learnability and operability of this keyboard, it can be seen in FIG. 38 that the letters forming the transfer coding group are presented in uppercase on the corresponding key buttons, while the non-transfer coding group is in lowercase. Or use different sizes, colors, and fonts to mark letter differences.

Operating this unit couldn't be easier. If the letter of the shift coding group is included in the text to be typed, the shift key must be pressed while pressing the corresponding key, and the letter will be presented in a non-discriminatory manner. Conversely, if it is necessary to key in a letter in the non-transition code group, pressing the corresponding key will make the letter appear in a different way.

According to the teachings of the present invention, the disambiguation function corresponding to this application can be physically integrated into the body of the phone. It can be used at the transmitting end of the communication and/or at the receiving end of the communication, such as a central computer in which the user communicates by telephone.

There are 4 keys on the standard telephone keypad that can be used to encode non-alphabetic information, such as mode change functions. The number of symbols used to encode non-alphabetic information on a telephone keypad can be further increased using the synthetic encoding symbology described above. Especially if less ambiguity is required, the number of shifted letters can be slightly increased. For example, if each of the 4 transfer keys is connected to one of the letter keys, although the number of keys required for each letter increases, the typing of characters can reach the point of completely disambiguating. To put it simply, there is a lot of room for imagination in the connection between the shift key and the symbol decoding subset. The internationalization of doctrine will be discussed more uniquely below, so far limited to the details of the English part.

It is worth mentioning that when the shift key and the remaining non-alphabetic keys arranged with *, # and 0 keys are used at the same time, at least 6 non-alphabetic keys can be encoded, such as punctuation marks, mode Toggle symbols and more.

Correct the error of using standard divergent coding. The keyboard of this application can still achieve what you want when you need to press the shift button but don't press it or press the shift button instead, especially when you are using it for beginners. Text encoding. Usually such operations will produce meaningless codes in cases where the disambiguation device thinks that the code sequence of the integrated harmonization/disambiguation code is correct. When the above situation occurs, instead of issuing a query, other disambiguation methods can be used regardless of whether the shift button is pressed, and the code sequence will be interpreted as if it were treated the same as the standard disambiguation code. Usually, such interpretations as above will find the text that the user originally intended to type.

It is worth noting that in the arrangement shown in FIG. 38, a high-relevance symbol (space bar) is paired with a low-relevance or non-relevance symbol (cancel key). This pairing basically asks the disambiguation software to correct mistakes like pressing the space bar instead of the cancel key or pressing the space bar instead of the cancel key.

Adapting to Tactile Type Query Method When querying is permitted as described above, it would be ideal to use the shift key as a scroll key. It is worth mentioning that, as long as there is a suitable software, it is possible to locate the function of this key button at any time whether it is shifting or scrolling. This key functions as a scroll when the device is in inquiry mode, and as a shift in other situations.

Alternate Positions of Shift Keys Referring again to Figure 38, we know that this application is designed to operate with existing standard telephones. If the manufacture of the phone is based on this application, then the key button 1005 should be added as a transfer function. Ideally, these additional keys should be placed on the sides of the phone so that they can be activated with the thumbs when the phone is held in the palm of the hand. Figure 38 shows the configuration described above. Other key buttons 10006 that can be activated by the fingers of the same hand (regardless of the left and right hands) can also function.

Low Frequency Character Screening in Queries Often a very high frequency character will differ from a very low frequency character. For example, in the case of CEHLNSTY in English, the very high frequency character "for" diverges from the very low frequency character "fop". Removing these very low frequency characters from the dictionary can effectively improve the lookup rate with minimal impact on said dictionary and the language it represents. For example, removing characters with a probability of less than 1 in 50,000 improves CEHLNSTY's query rate to one query per 46 characters. Such screening can be achieved using methods such as "gap factors". A "gap factor" is the ratio between two characters being queried, eg, the ratio of the highest and lowest frequency characters. For example, assuming that the gap factor is 500, we will obtain the distribution shown in Figure xxx, and the two codes specifically pointed out in the figure are standard divergent codes (SAC) and CEHLNSTY codes.

Internationalization There are two main keys to internationalization for this application, and users familiar with the teachings of this invention can easily solve it. These are 1) the manipulation of intonation, and 2) the creation of inductive codes that can be used in multiple languages simultaneously. These two points will be briefly discussed later.

The handling of accents. Many languages use letters with circumflex accents. For example, "e" in French can be written as "e", "è" or "é". Without circumflex, "èlève" (meaning "student") would diverge from "èlevé" (meaning "to rise"). We can start with the use of another shift key. When this transfer key and coding key are used simultaneously, there will be the function of selecting letters containing circumflex. This key can be called circumflex transfer key. For example, in French using the CEHLNSTY encoding, the circumflex shift key can be used in conjunction with the "def" key to encode "è" or "e", and then rely on the disambiguation function to determine which is the appropriate circumflex accent. Using the above steps, we found in a group of French character frequency data in terms of CEHLNSTY coded (search, query) rate, it is (38, 3) when not using the circumflex transfer key, and using the circumflex transfer key When the button is pressed, it is (584, 24). The bottom row of keys on the telephone keypad can all be used for circumflex shifting. On special keypads, additional keys can be added to provide circumflex shifting. From the point of view of ergonomics, it would be ideal if the operation of the tone shift key can be similar to the normal shift key. For example, one direction of the thumb movement can be used to code the operation of the common transfer key, and another direction of the thumb movement can be used to encode the operation of the accent transfer key.

Multilingual Differing Encodings Since the data are often different between languages, an encoding optimized for one language may not necessarily be optimized for other languages. An encoding that is high-touch typing for one language may not be suitable for another language.

For example, CEHLNSTY optimized for English performs worse when used in French than an encoding chosen specifically for French optimization. While CEHLNSTY is high touch for French typing, this is not the case for every language.

In order to achieve an economical advantage, the manufacturer should produce a machine that can be used in various language regions. For typed devices such as mobile phones, the key buttons will be marked with different codes designed for them. If a code that is applicable to multiple languages is selected, the above key button labeling method can be used on all production lines regardless of any language area.

Using the techniques described above, it is possible to program divergent codes that are optimized for a variety of different languages at the same time. In the multilingual optimization method, the step of determining the proportion of ergonomic standards can be regarded as one of the substeps of the step of determining the proportion of multilingual standards. Different specific gravity methods must be used in different situations. For example, when optimizing for both English and German data, perhaps the weighting of encodings in English is more important than in German.

Maximizing the minimum performance would be the ideal weighting method. The steps above are called the lowest-highest steps.

Assume that a set of different languages 11, 12, ..., 1n and a set of ergonomic criteria e1, e2, ..., em are to be optimized. Given the discrepancy codes c1, c2 and the use of em as an ergonomic criterion, if the minimum value of em divided by language 1n is higher for c1 than for c2, then c1 has a higher rating than c2. When there are many ergonomic criteria that must be optimized, it is possible to find that a code is particularly good for one particular ergonomic criterion and poor for another. Taking this example as an example, ergonomic standards must be considered in proportion to the importance of each other as mentioned above.

According to the above concepts, for example, assume that there is a group of languages that need to optimize the search error rate and query rate. In this example, we will use 8 standard typing functions, one random typing function, and the random typing function of the accent shift key in an alphabetical order. Therefore, these 8 kinds of standard key-in functions can be activated simultaneously with one of the random-type key-in functions as mentioned above to achieve the application of comprehensive coordination/difference coding.

First, the language group that needs to be considered for optimization includes French, Italian, Portuguese, and Spanish, and each language has its own reference data.

The following codes are easily obtained using guided random progression: joz m b h x akn r pw d iy l gq t ev c fu s. This encoding is respectively (3250, 265), (11400, 3800), (4720, 505) and (6280, 400) to French, Italian, Portuguese and Spanish (search error rate, query rate). However, this code performs relatively poorly on Dutch, English and German, which are (65, 4.8), (93, 10) and (360, 13) respectively.

The following codes can be obtained when optimizing Dutch, English and German in the same amount of time spent on a calculator: cjk r biy l fv e mo a sz phx g tu d qw n. And (search error rate, query rate) are (1220, 44), (816, 44) and (480, 47) respectively. This encoding is respectively (253,20), (306,50), (525,36) and (4236,272) to French, Italian, Portuguese and Spanish (search error rate, query rate). Although the above results are good for these randomly sampled languages, they are still far from the results optimized specifically for language diversity. The above results indicate that the greater the differences between each language, the lower the commonality of encoding representation.

In reality, whether or not to include certain languages in a multilingual optimization plan is a business rather than a conceptual decision. The point of this invention is that even with minimal performance in the language of choice, the code still needs to be capable of high-touch typing. In the previously examined cases, optimizations for French, Italian, Portuguese, and Spanish produced high-touch codes rated C, but minimally high for Dutch, English, and German. The performance is only A grade high touch key.

One-handed high-touch palm-type device can be seen in the above-mentioned integrated harmonized/discriminated coding application, and the superior keying function can be used for divergent coding symbol coding to reduce the divergence of the overall system. The above existing application shows that the same set of keying functions can be used to coordinate the encoding of information and divergent encoding symbols. The diverging codes in the above example can be represented using multilevel codes: the keying sequence in the first segment is used to select the first subset of codes, the keying sequence in the second segment is used to select the second subset of codes, etc. wait. Ideally, the second subset is a subset of the first subset, and the third subset is a subset of the second subset (that is, a subset of the first subset) . This is the "point occupation" method familiar to insiders. So far, it is not fully understood that a) the number of consecutive sub-separations in a set of symbols can be limited by the fact that the smallest subset contains more than one symbol, becoming a divergent code. b) Minimize the ambiguity of the final ambiguity encoding by selecting the sub-segmentation properties. c) Disambiguation can be optimized while at the same time improving other ergonomic criteria eg while following tradition. d) The transfer of the level on the level can be achieved by simply pressing the matching button under coordination.

Specific extended arguments for several of the above findings will be discussed in more detail with reference to Figures 39 to 47. Notably, this is one of an infinite number of devices that can be made according to the teachings of this invention. All keying devices that use the dotted occupation method to create codes and at the same time improve other ergonomic criteria such as optimization while following tradition are included in the above scope of application.

This application device has a high tactility that can be used with one hand. Its additional superior features include:

1 Availability of fully non-discriminative text typing.

2. Optimize the usability of the high-touch and divergent text typing method.

3 Availability of the minimum keystroke mode when capturing data.

4 and the above three modes have the highest ergonomic compatibility with each other.

Ideally, in addition to the above ergonomic criteria, scanning time should also be one of the criteria to be optimized. Optimization of scan time is discussed in the next paragraph.

In order to establish a tactile key device based on multi-level differential coding, we will discuss in an overview with reference to FIG. 39 .

In a first step 150, a set of decoding symbols of the second level must be selected. These decoding symbols are represented by divergent codes, which may also include letters such as a to z in English.

In the next step 151, an ergonomic criterion suitable for all multilevel codes must be selected. This ergonomic criterion can be, for example, high tactility or finding errors. Often many of the ergonomic criteria applicable to all multilevel codes can be applied simultaneously. In the next step 152, the selected second-level decoding symbols are divided into subsets. Each sub-set of the second level is assigned a decoding symbol, so that the overall encoding can be optimized according to the selected ergonomics. This approach does not differ from any least-disambiguation encoding method so far. However, there may be additional restrictions, such as restrictions on the number of encoding symbols that can be allowed, which also enable the implementation of the next step of the method. In this next step 153, the second level coded symbols will be collected into groups. These groups are considered as decoded symbols for the first level of encoding. In the absence of accidents, the encoded symbols of the second-level encoding will become the decoding symbols of the first-level encoding. In this way, first-level coding symbols are assigned to each group, which also forms a first-level divergent coding. Other optimizations for ergonomic criteria can be made in the designation of the second-level coded entry groups. Roughly speaking, each level of multi-level coding can be optimized according to different ergonomic criteria. These criteria may be the same as or different from the ergonomic criteria used for all multilevel code optimizations. In a final step 154, this multi-level encoding is applied to a typing device.

In the description of this multi-level encoding method, the second-level encoding method is to advance this method to the first-level encoding. When operating a multi-level encoding device, this sequence will be reversed: first, a component of the first level of encoding is selected by operating the keying device, and then the second level is selected by further operating the keying device One of the components of the code. This can be said to be the essence of this point occupation method. This method can also be applied to third and higher level codes, which will be obvious to the skilled user.

In practical applications, the properties of each level of multi-level coding must be optimized simultaneously to achieve all the desired properties of multi-level coding. This current application can concretely illustrate how to plan and execute this synchronization optimization. In short, we can illustrate this architectural approach in this application through Figure 40. In order to be able to fully display this procedure, we selected three ergonomic standards for all multi-level codes, the first level codes set two standards, and the second level codes set three standards, thus forming this Multi-level encoding. In this application, the three ergonomic criteria applied to multi-level coding are ease of touch, search rate, and search error rate. The first level of coding is optimized for structural accuracy and alphabetical order, while the second level of coding is optimized for separation equality, structural accuracy and substantial English alphabetical order.

The first step 3100 in this application is the selection of (second stage) decoded symbols. That is to choose from the letters a to z. Next, in steps 3101, 3102 and 3103, respectively select the easy-to-touch key degree, query rate and search error rate as ergonomic standards in the multi-level coding. Next, structural accuracy is selected in step 3104 as an ergonomic criterion. Because this device is designed for being able to type with fingers while holding the device, the structural accuracy reaches the highest point here, that is, each of the 4 typing devices and the 4 corresponding first-level coding symbols is composed of a different Fingers are in charge.

In a second level step 3105 structural accuracy is selected as an ergonomic criterion. Each encoding symbol in the first-level encoding will correspond to several second-level encoding symbols. If each of the 4 first-level symbols corresponds to 4 second-level coding symbols, the structural accuracy of the second-level coding can reach the highest point, that is, in the case of 16 second-level coding symbols Structural precision is at its peak. If 26 second-level decoding symbols are distributed among the second-level decoding symbols, and 1 or 2 of the second-level decoding symbols are connected to each of the 16 second-level encoding symbols , then, the 16 second-level encoding symbols can also be connected with the second-level decoding symbols, thereby achieving the best separation equality. This distribution then means that between 4 and 8 second-level decoded symbols will eventually be connected to each of the 4 first-level codes.

Next, in step 3106, we selected English alphabetical order as the ergonomic standard in the first-level coding. Optimizing in accordance with this standard requires synchronous optimization of both the first and second level codes. That is, the letters a to z need to be arranged in alphabetical order to be displayed on the display corresponding to the input device connected to each finger. Since these displays are arranged according to the order of the fingers, this arrangement means that the first part of the letters in the alphabetical order must be connected with the first level of coding symbols connected with the input device connected to the first finger. Level decoding symbols are connected. Equally, the second group of letters, the alphabetic order of the continuation of the second group of letters, must be connected with the second-level coding symbols connected with the first-level coding symbols connected with the input device connected to the next finger, and the other two first-level coding symbols are connected. The first-level coding symbols can also be deduced in the same way. Optimizing by alphabetical order therefore corresponds to choosing a permutation cell for the 26 letters, in the same manner as discussed in other applications of this invention. This time, each of the 4 components of the grid must have 4 or 8 sub-sections, so that all the ergonomic criteria we list can be optimized simultaneously. This will be the same as shown in the specification of the best mode for this application, that is, when optimizing according to all other ergonomic criteria considered, we can find out Possibly equal-separated encodings, the same for first-level encodings.

Finally, in step 3107, we selected the substantive English alphabetical order as the ergonomic standard in the second-level divergent coding. This means that we will, as far as possible, arrange the letters in alphabetical order, given that all other restrictions on the assignment of letters to second-level coded symbols are known. The degree of deviation from strict alphabetical order can be measured in many ways, for example, by the number of paired permutations required to bring a known permutation into strict alphabetical order.

Referring now to Figures 41-47, we will illustrate an easy-touch typing device, which can be typed with one hand, encodes at least the letters [a-z] and includes a code structured according to the method described above . To make the device touch-friendly according to the method of this application, it is necessary to fix the separation of symbols into subsets, subsets of subsets, etc., that is, for example, according to which symbol was previously typed. No change. This fixed separation is only applicable to the required degree of easy touch, but the principle of this invention can be applied in a wider range. For example, using the character completion mechanism can greatly reduce the number of keystrokes. The behavior of the character integrity mechanism is complex and difficult to predict, and the device with the character integrity function has no possibility of being touch-sensitive. However, in the case that the lower the degree of ambiguity, the better the character integrity can be improved, and the same optimization procedure for easy-touch coding can achieve an effective character integrity mechanism. It can be seen that adding a character integrity mechanism to an easy-touch typing device does not make the device outside the scope of this invention.

Now we make further restrictions on this application, that is, the part of the device that can be typed must be able to be held by one hand, and it must only be typed with the hand that holds the device. . To limit the need for numerical actions, most symbols can be typed by a series of operations on the following 5 keying devices: 4 keying devices operable by the fingers of the hand holding devices 2100-2103 And a keying device that can be operated by the thumb of the hand holding the device 2104. The device shown in Figure 41 is designed to be held by the left hand; obviously, a symmetrical right-handed device could also be designed, or an ambidextrous device that can be operated with one hand.

Ideally, associated with each of the keying devices 2100-2103 is a visual display 2106-2109 showing the subset of components currently associated with known keying devices. The keying device is operated to select the corresponding subset. Key-in device 2104 may be used to further refine subset selection and/or to select subsets of other symbols. For example, the single symbol "space" may be associated with the input device 2104; this or other symbols associated with the input device 2104 may be preferentially displayed on the display 2110. The letters [a-z] can be distributed among the 4 keying devices 2100-2103. Preferring such a keying device letter distribution minimizes the degree of ambiguity (finding error rate and/or query rate) and at the same time respects the traditional alphabetical order. Such compliance can help novice users to find the desired letter simply by scanning the candidate letters easily.

Figure 42 illustrates an arrangement of letters [a-z] in which letters [a-f] are associated with the first keying device 2100, [g-l] are connected with the second keying device 2101, and [m-r] are connected with the third keying device 2102 Finally, [s-z] is connected to the fourth input device 2103. These associations constitute the first-level subsets in the first-level coding. In short, ideally, 4-8 letters would be chosen to be associated with each of the 4 keying devices: the subset of letters associated with each keying device could be further subdivided into 4 subsets , and each of the four subsets cannot contain more than two letters. The function of this restriction will soon become apparent, and it will be obvious to a skilled user how to extend the application of this principle to languages with different numbers of symbols and different numbers of keying devices .

FIG. 43 illustrates an example set of second-level subsets separating the first-level subsets shown in FIG. 42. FIG. Figure 43 is a table with four columns and four columns. The columns are marked by the keying devices enabled in the first step, and the columns are marked by the symbols associated with each keying device in the second step. Therefore, for example, if the input device 2100 is activated at first, then in the second step, the symbol ac will be connected to the input device 2100, and the symbol be will be connected to the input device 2101 and so on. In cases where the limits on subset size and lookup rate are as explained above, choosing this arrangement minimizes the lookup error rate and lookup rate. Using our reference data, the lookup error rate and query rate of this encoding are (1100, 69). It is worth noting very carefully that in this example, the letters in the first-level subsets are arranged alphabetically, but the letters in the second-level subsets can only be partially arranged in alphabetical order. For this example, we have decided to relax the alphabetical order constraints at the second level, thus improving lookup error rates and lookup rates, and producing an encoding that is as touch-friendly as possible. This shows that alphabetical order is and/or is not optimizable, just like any other ergonomic criterion, and that the properties of optimization can be different at each level in a multi-level divergent code. Again, the advantage of the English alphabetical order is that the scanning time is shortened, especially for novice users. Due to the small number of symbols displayed in the second level, the scan time is already short, and can be further shortened by the mechanism we are going to discuss now.

To type a desired letter, first, the user activates one of the keying devices 2100-2103 corresponding to the first subset containing the desired letter. Next, the user selects one of the second-level subsets again by activating one of the input devices 2100-2103 corresponding to the first subset containing the desired letter. Figure 44 illustrates an example of the operation of such a device, where the user types the letter e. Referring to FIG. 42, we can see that the letter e is connected with the input device 2100 through the second-level mode. When the user activates this input device, the display will be as shown in Figure 44. The letter e is now connected to the input device 2101 . When the input device is operated, the letter e is output. The same operating sequence of the input device can be used for the selection of the letter b, so that this coding is different. As in other applications, whether the letter b or e is desired will depend on its context through a disambiguation mechanism.

Characters are typed through sequentially selected letters according to this method, and the character is terminated by activating the typing device 2104 connected to the thumb and both hands. It is worth noting that while a two-key method is used to encode each letter, this keying device forms the basis for a one/two-handed application. More specifically, if one hand is used to mark the first keystroke of each letter, and the other hand is used to mark the second keystroke of each letter, then the first and second The messages of the two keystrokes can be entered synchronously. Many practical applications can also be based on this. For example, the "fingering" mechanism in [4] can be used as a practical basis for a one-handed/two-handed application. The encoding proposed in [4] is based on the fact that the motion sensor can sense several different positions of each finger to disambiguate each letter. This would require a more accurate sensor. However, simpler sensors can be used if a two-handed form is used for this application. These sensors only need to record binary (up/down) information for each finger. The complexity of software and hardware can be reduced at the same time. Additionally, devices built according to the principles of this invention will be relatively easy for users to learn and operate.

Visual Cache Scanning time is the time to visually find a desired letter from a set of letters. Typists who do not follow the normal typing method will visually scan the keyboard to find the next letter, and then press its corresponding key button. Scan time depends on several factors, including the user's familiarity with the keyboard layout. Typists who do not follow formal typography may roughly know where the desired letter is, and only visually scan to confirm, or find the correct location. Since the scan time improvement is achieved through the average user's familiarity with the English alphabet, this application chose the English alphabet as the first level of encoding. In yet another alphabetical arrangement, certain specific letters are selected from a group of letters displayed on a given key in a selected area for clear distinction on a visual display associated with the key. was chosen. These letters are the ones most likely to be chosen in any case, and placing them in a clearly identifiable position makes them easier to find. The principle is similar to the cache used in some computer processors to store recently used data in registers so that they can be quickly retrieved next time they are needed, based on the fact that recently used data will be more likely to be used again under the hypothetical situation. Here, letters are cached not because they have been used recently, but because of the probability that they will be used next time, given that the relevant information about the language is known. However, the term visual cache is still applicable.

We will now illustrate an application of visual caching in context of this application. It is worth noting that this invention allows a wider range of adjustments without having to adjust its main characteristics, for example, in the size and location of the cache, how to place the cache, how to label the cache, etc. .

From our standard data analysis, we found that for the letters [a-f] connected to the keying device 2100 through the first level encoding, "a" is the first letter of a character most likely to be. Similarly, among the letters [g-l] connected to the input device 2101, "i" is the first letter that is most likely to become a character, and "o" is the most likely among the letters [m-r] connected to the input device 2102. and "t" is the most probable one among the letters [s-z] associated with the input device 2103.

By arranging the letters a, i, o, t in an obvious position on the screen, for example, the upper left corner of the screen where each keying device is connected. This will make these letters the first letters of contact on a standard left-to-right, top-to-bottom visual scan with each connected screen. Ideally, other single-letter choices other than this alphabetical order, such as reserved alphabetical orderings for other letters in the subset. The distinction between letters in the cache and other letters can be further marked by choosing a different font color, size, style, etc. for the cached letters than for other letters.

Referring now to Figures 45 and 46, we will see that this discovery can be used to reduce scan time. Figure 45 illustrates how the character "think" is typed without using visual caching, while Figure 46 illustrates the same character with visual caching. Thus, in FIG. 45, the letter "t" is typed by first activating the corresponding letter "t" of the first keying device 2103. Before the first input device is activated, the screen display will be as shown in the second column of this figure. When the 2103 is enabled, the screen display will change to that shown in the third column. When the input device 2101 is activated, the letter "t" is its output result. When other letters of the character "think" are typed, the screen display will have a similar change.

In FIG. 46 , identifying whether a letter is stored in the visual cache is indicated by representing a letter stored in the visual cache with an uppercase letter, while a letter not stored in the visual cache is represented by a lowercase letter. Based on our reference data, we found that 42 percent of characters do not start with a, i, o, or t. Thus, 42 percent of the time the user will find the desired letter in the cache just in time when they start typing a character.

When a character is typed, the most likely next letter is changed in the context of typing that character. It can be seen that the letters selected to be stored in the visual cache will change according to the characters typed, and also depend on which character is typed.

In the example of the character "think", the letter "t" can be found in the visual cache before and after the activation of the first keying device and before the activation of the second keying device, as shown in the first four columns in FIG. As shown, each column corresponds to a keying device display. Once the letter "t" is selected, the letters in the visual cache become a, h, o, w, as shown in Figure 46 in the second set of four columns. After selecting the first keying device (keying device 2101) as starting to key in the letter "h", according to the reference data, there are only two possibilities for the letter that will form part of the character again, that is, the letter "h" and the letter "i" , both of which will be in the visual cache in the second display. Continue to carry out letters i, n, and k in the same way, and we will find that the required letter will always be in the visual cache of this character. Indeed, after activating the first keying device to key in the letter "i", there are only two letters that may be needed if the user actually keyed in the characters in the information library. Then, in the example, when the second input device is virtually inactive, the letter "i" can be output immediately after the first input device is activated.

Thorough disambiguation programs, as well as other symbolic typing, generally provide a completely non-disambiguating method for symbolic typing on a disambiguatingly encoded typing device, as already explained. In the present application, a simple method of providing non-discriminative typing is by providing an additional non-discriminative typing device 2105 as shown in FIG. . However, other locations can also be chosen.

In this application we have chosen to limit the size of the second level subset to at most 2 symbols in order to limit it. Thus, the disambiguation mechanism will always be able to correctly select the desired letter, or it will always be able to select another incorrect but paired letter. Any disambiguation softkey may signal the two symbols it will select when its corresponding keying device is selected by the user. This signal can be used to provide feedback to the user, for example, by highlighting selected letters. If the selected letter is not the required letter, then the user has the opportunity to choose to enable the complete disambiguation device 2105 in Figure 42 to force another non-highlighted symbol to be selected.

Shown in FIG. 48 is an example of using this non-discriminative text input device. And in Fig. 46 and 47, then illustrate how this character of " think " is keyed in. Here, when the non-discriminative keying device 2105 is activated after the first and second keying devices for typing the letter of the character "think" are activated, what is displayed in the fourth column is the letter to be output. For example, suppose we first activate the input device 2103 and then activate the input device 2101 to type the character "think", then further activation of the input device 2105 will select the letter "u". Then the letter "u" will be the first letter of this character. All possible letters can be typed in this method for disambiguation. When only one letter is included in this second-level subset, this letter can be typed without activating the input device 2105 for disambiguation, that is to say, the input device 2105 is not suitable for this purpose. The letters d, f, h, l, and p are always typed disambiguated according to the specified encoding.

Accessibility: Measuring and thresholding the accessibility of keys is a new and innovative concept to illustrate the definite specification of devices. The breadth of this concept has been illustrated by a series of applications, that is, by pushing its limits to show its range of application.

To add further certainty to the account of reachability, this chapter will propose an additional numerical property of reachability that will make the reachability of any divergent code measurable, whereby It can be decided whether this code belongs to the scope of this invention.

Linguistic Data As we mentioned before, representing a language in terms of a language corpus is a research topic for linguists. In the interests of numerical certainty, we define a representative language corpus as a total collection of at least 10 million characters randomly selected from general tabloids in the target language.

Number of buttons We need to define the following four kinds of button numbers: actual button number, coordination button number, effective button number and hybrid effective button number. Actual number of keys: the number of keys used for symbol encoding. A basic qwerty keyboard has 26 buttons each labeled with a letter, a shift key, and a space bar, so the actual number of buttons is 28. Number of coordinated keys: The number of specific combinations of keys used for symbol encoding. Taking the most basic qwerty keyboard as an example, the shift key can be combined with any letter key to form a capital letter, that is to say, the number of coordination keys of this keyboard is 28+26+-1=53. The turn key cannot encode any symbols. Without exception, a keyboard identical to a qwerty keyboard would consist of 53 physical keys, each of which encodes a single symbol, whether an uppercase letter or a lowercase letter. Indeed, some early typewriters were constructed in this way.

Number of effective buttons: When a group of symbols that can be represented by p in a different code, a group of language materials, and several actual buttons, one has the lowest possible search error rate and query rate An optimal ambiguity code exists if this is the only constraint on the code. Let us use pl and pq to represent these two data respectively. Any divergent code on any number of physical buttons will have an effective button number p when its lookup error rate and lookup rate are equal to pl and pq. It is impossible for a keyboard with fewer than p physical buttons to support a divergent code with an effective number of buttons equal to or greater than p. It is quite possible, and usually so, for a divergent code on a real key number p that the effective number of keys for a divergent code on a physical key is less than p. Number of Mixed Active Buttons: This is an experimental finding that the lookup error rate and lookup rate of substantially optimal divergent codes are substantially linked by a power law, for example, through the experimental results in Figure 11 . The description here is mainly in English, and the logarithm of the substantially optimal query rate is directly proportional to the logarithm of the substantially optimal search error rate.

We can this find a single number that defines the link between the lookup error rate and the lookup rate of an encoding: the position of the encoding (lookup error rate, lookup rate) is indicated on the log-number best link.

For example, consider standard divergence encoding. The position of this encoding (lookup error rate, lookup rate) is (29, 2.2). The linear intersection of this position is marked on the best connection line of Fig. 11, we can find out this numerical value of 5.96, just the mixing effective key button number of this standard divergent coding. Although the standard divergence code is defined on 8 physical keys (the number of coordination keys is also 8, because no coordination is needed here), it is the best in terms of divergence with a 5.96 physical (or coordination) key number Encodings are equal. It is true that there are a small number of physical buttons that are not practically feasible, but these results show that it is possible to find a 6-button physical optimal code that has a better lookup error rate and lookup rate than the standard divergent code .

These considerations allow us to define an exact, albeit out of context, numerical threshold for the substantive optimum of mixed lookup error rates and lookup rates: if its mixed effective key count is within 0.01 of its coordinating key count Within, after reference to these data, a code can be said to be substantially optimal in the sense that no other ergonomic constraints are imposed on the system. We can also define an exact, albeit out of context, threshold of tactility: an English ambiguous code can be defined as having the highest tactility when the number of valid keys in the mix is at least 10. We can extend this definition to other languages by restricting that a code must be touch-friendly, and its lookup error rate and query rate must be higher than or equal to an English touch-key code. However, the number of effective key buttons mixed with the standard divergent coding is less than 10, so it will not be an easy-touch key according to various considerations in this chapter.

By measuring the number of effective keys in the mix, any divergent codes will be filtered with the attribute of easy-to-touch keys. As an example, the complex coordinating/discriminating code discussed above: ab c df e gi h jk l mo n pqr s uv t wxz y has 9 real keys: 8 on the keypad of a standard telephone Plus a shift key. The number of its coordination keys is 16; that is to say, it is equal to the differential coding of 16 independent keys minus one transfer key. Without applying a spacing factor, the (lookup, lookup) rate would be (431, 21) corresponding to a mixed active button number of 12.8. Considering that the interval factor is 500, this data will be increased to (440, 46), and corresponding to it is a mixed effective key button number of 13.75. This code will be easy to touch with or without spacing considerations. It is worth noting that the number of effective mixing keys here is less than the number of coordination keys, but this encoding is actually optimized because there is an additional restriction on the arrangement of English letters. Equally, the application of a one-handed synthetic coordination/difference coding will have a (search, query) rate of (1100, 69), and a resulting mixed effective key number is 15 codes, although its actual key The number is four and the number of coordination keys is 16. This is an easy-to-touch code, and the difference between the number of coordinated buttons and the number of mixed effective buttons is created by the additional ergonomic constraints of the alphabetical order of the first-level codes in English and Chinese. Taking this additional constraint into account, the encoding is substantially optimized.

In contrast with this, Fujitsu's 14-substantial key-button code, pn gt cr zk wj a ehi so ud xf ym vl qb, its (searching, query) rate is (105, 4), and the effective key-button number of mixing is 8.47. This code is neither substantially optimized nor touch-friendly, although the actual number of buttons is greater than 10.

Although examples of applications of the invention have been described herein with reference to the accompanying drawings, it should be understood that the invention is not limited thereto, but many others can be made by a skilled user without departing from the scope of the invention. Changes and adjustments may affect it.

references

[1] http: /www,fujit su.co.jp/hypertext/news/1996/Jun/ms-txt.htmlhttp:/www,fujit su.co.jp/hypertext/news/1998/May/27-e .html http: /www,fujit su.co.jp/hypertext/news/1996/Jul/text.htm

[2] US5,818,437, Shrinking Keyboard Dissimilarity Calculator. US5818437 October 6, 1998

[3] Reducing keyboard ambiguity system, PCT/US98/01307; WO98/33111

[4] "Subject-integrated fingering": Wireless Wearable Belt Keyboard, CHI 97 Electronic Publications: Papers, Authors: FUKUMOTO, Masaaki and TONOMURA, YoshinobuNTT Human Interface Laboratory, 1-1 Hikari-no-oka.Yokosuka-shiKanagawa -ken, 239 Japan (JAPAN) online reference data:

http://www.acm.org/turing/sigs/sigchi/chi97/proceedings/paper/fkm.htm#U21.

Claims (46)

1, a kind of device of keying in is characterized in that, includes:

The key entry mechanism that user's operation is reacted is to generate the coded identification sequence;

The different coding mechanism of easy touching formula that the coded identification sequence is mapped across the decoding symbol sebolic addressing; And

The output mechanism of a selectivity output decoding symbol sebolic addressing when the reaction user operates described key entry mechanism.

2, the device of keying in as claimed in claim 1, it is characterized in that: the different coding mechanism of described easy touching formula substantially physically is away from described key entry mechanism, and further comprise a communicator with the signal and communication that generates by described key entry mechanism to described different coding mechanism, and this signal is corresponding with described coded identification sequence and be mapped across the decoding symbol sebolic addressing by described different coding mechanism.

3, the device of keying in as claimed in claim 1, it is characterized in that: described easy touching formula difference is encoded to an optimized different coding, its according to comprise the inquiry rate at least, search error rate, structural precision, outward appearance accuracy, separation structure, tradition reservation degree, cross-platform compatibility, the regularity of arrangement and one ergonomics standard in sweep time carry out optimization.

4, the device of keying in as claimed in claim 3 is characterized in that: described optimized difference be encoded to come down to best.

5, the device of keying in as claimed in claim 3, it is characterized in that: one in the described optimized ergonomics standard meets traditional, the wherein said tradition that meets is that the standard that expression and alphanumeric tokens and standard telephone key button are keyed between the function is got in touch relation, and the standard contact that this alphanumeric tokens and standard telephone key button are keyed between the function concerns first group that comprises non-letter character and 4 non-alphabetical key buttons, and second group the relation of getting in touch of letter character and 8 letters and non-alphabetical key button.

6, the device of keying in as claimed in claim 5, it is characterized in that: having wherein an optimized ergonomics standard at least is from by the inquiry rate, search mistake, select in the group that structure accuracy and outward appearance accuracy constitute, this can be keyed in device and also comprise at least one harmony key entry device, it is complementary key button that this at least one harmony is keyed in device, device and 0 keyed in described non-letter, 1, * with the conventional contact of symbol such as #, this harmony is keyed in device and the key button of selecting that interrelates with letter character simultaneously or operation in regular turn from above-mentioned second group, and and the subclass of the letter character that interrelates of described key button also interrelate with the coded identification of described different coding.

7, the device of keying in as claimed in claim 6, it is characterized in that: one operation in described letter and the non-alphabetical key button interrelates the subclass and the coded identification of corresponding letters individually, interrelate with described different another coded identification of encoding and key in the operation of identical letter that device combines and the non-alphabetical key button subclass that other are alphabetical with described at least one harmony, one in this other alphabetical subclass and described letter and the non-alphabetical input unit interrelates.

8, the device of keying in as claimed in claim 7 is characterized in that: the operation that at least one harmony that has one of them described letter and a non-alphabetical key button is keyed in device makes in the one group of single letter that comes from the corresponding letter one interrelate with corresponding coded identification.

9, the device of keying in as claimed in claim 8 is characterized in that: described one group of single letter is respectively C, E, and H, L, N, S, T and Y key in device label from 2 to 9 corresponding to described letter and non-letter.

10, the device of keying in as claimed in claim 1, it is characterized in that: also include first subclass of main key button and second subclass of less important key button, wherein said difference is encoded to multistage different coding, activate first subclass of selecting the decoding symbol one the first time in the described less important key button, activates second subclass of then selecting the decoding symbol for the second time.

11, a kind of device of keying in, include a plurality of key buttons, these key buttons contain be useful on a plurality of symbols in tool importance on the statistics of coding a plurality of on statistics the symbol key button of tool importance, described a plurality of radix at the symbol key button that has importance on the statistics is less than described a plurality of radixes of the symbol of importance of having on statistics, therefore described at least a plurality of on statistics one in the symbol key button of tool importance with corresponding more than one the described symbol of tool importance on statistics, and the selection of this relative response to make the described device of keying in be high touching degree.

12, the device of keying in as claimed in claim 11, it is characterized in that: described relative response satisfy include the inquiry rate, search mistake, structure accuracy, outward appearance accuracy and separate in the ergonomics standard of structure at least one, described relative response and traditional consistency are optimized basically.

13, the device of keying in as claimed in claim 1, it is characterized in that: with respect to reach the basic statistics of extracting the text collected works of this language of representing from including at least one ten million word, the error rate of searching that the different coding mechanism of high touching degree has the inquiry rate of at least 10 words of each inquiry and searches at least 100 words at every turn.

14, the device of keying in as claimed in claim 1 is characterized in that: the described mechanism that keys in has the seemingly arrangement of qwerty formula.

15, the device of keying in as claimed in claim 1 is characterized in that: described key entry mechanism is embedded on the steering wheel of the vehicles, thereby the described device of keying in can be used when driving by the driver.

16, the device of keying in as claimed in claim 1 is characterized in that: described easy touching formula coding has the arrangement of basic English alphabet.

17, the device of keying in as claimed in claim 1 is characterized in that: described easy touching formula coding is simultaneously for still being high touching degree under the situation of more than one language.

18, the device of keying in as claimed in claim 1 is characterized in that: described easy touching formula coding is for the key entry at given number, have an inquiry rate at least and search error rate the best may encode 5% in.

19, the device of keying in as claimed in claim 1 is characterized in that: also include Touch Screen, this Touch Screen includes described input mechanism, then is optionally transparent with the decoding symbol that is associated of each described key button.

20, can key in device according to claim 1, it is characterized in that: also comprise a mechanism so that described device is connected with position on calculator screen the position in the plane, and become the function of position on the described computer screen with the decoding symbol that is associated of each described key button.

21, the device of keying in as claimed in claim 20, it is characterized in that: also comprise key button, handle and the display screen that can activate with thumb, it is effective to the no different symbol of input that described thumb activates key button, can allow the described device of keying in move along the plane when hand is in a described handle during for the comfortable position of typewriting action, and described display is effective to showing a group code that interrelates with current key button by the palm applied pressure.

22, the device of keying in as claimed in claim 1 is characterized in that: also comprise four substantially the same sizes, flat part basically, each all is to be connected with each other collapsibly partly, and the described device of keying in is two foldings.

23, the device of keying in as claimed in claim 1, it is characterized in that: a described son group keying in mechanism is radix 13, the similar mechanism coding of described high touching formula videos two letters among the a-z to each subclass of described key button at least, and the described device of keying in can only be typewrited can key in the identical posture of device with the no difference of typewriting with both hands basically with singlehanded.

24, the device of keying in as claimed in claim 23 is characterized in that: described decoding symbol is that described distribution allows an optimized in appearance basically layout to the distribution of described key button.

25, the device of keying in as claimed in claim 1 is characterized in that: described key button distributes according to standard digital keypad pattern.

26, the device of keying in as claimed in claim 1 is characterized in that: the different coding mechanism of described easy touching formula is a multistage different coding mechanism.

27, the device of keying in as claimed in claim 26, it is characterized in that: also comprise one group of main key button, described multistage different coding mechanism comprises the different coding of a first order and the different coding in the second level, a coded identification of extracting is selected in one first operation in the wherein said main key button from the different coding of the described first order, and a coded identification of extracting is selected in one second operation in the described main input unit from the different coding in the described second level.

28, the device of keying in as claimed in claim 27, it is characterized in that: describedly key in the device keyed in that device is a hand-hold type, described one group of main key button is a radix 4, and therefore described main key button is corresponding one to one with the finger that the user holds described this hand of device.

29, the device of keying in as claimed in claim 28, it is characterized in that: also comprise a group display equipment, one in this group display equipment each and the described main key button interrelates, the decoding symbol of the different coding in the described first order of each the operated in demonstration in the described display and the second level.

30, the device of keying in as claimed in claim 29, it is characterized in that: described multistage different coding mechanism is with respect to search error rate and the inquiry rate is optimized, described first order coding is to carry out optimizedly with respect to the English alphabet order, and described second level coding is with basic English alphabet order and separates equality and carry out optimized.

31, the device of keying in as claimed in claim 30 is characterized in that: also comprise a no different key button, the different key button of described nothing can be used to select clearly in the decoding symbol of the different coding in the described second level.

32, the device of keying in as claimed in claim 29, it is characterized in that: at least one most possible decoding symbol at first is docile and obedient preface and is presented at least one described display and reduces sweep time, and described at least one most possible symbol can distinguish in other less possible decoding symbol obviously.

33, the device of keying in as claimed in claim 26 is characterized in that: also comprise one group of less important key button, described less important key button can be used to generate the symbol of being encoded by no differently.

34, a kind of manufacturing can be keyed in the method for device, comprises the steps:

Be chosen in the multistage different coding as the decoding symbol of representative, this multistage different coding comprises the different coding of a first order and different coding the in the second level;

Select an ergonomics standard so that this multistage different coding is carried out optimization;

Described selecteed decoding symbol branch is gone into second level subclass;

Distribute a second level coded identification to give each second level subclass, therefore making selected ergonomics standard is optimization;

Described second level coded identification is grouped together, thus the second ergonomics standard, may be identical with the optimized ergonomics standard that is used for multilevel coding, be optimization;

Distribute a first order coded identification to give each described group, the multistage different coding of therefore relevant with second level coded identification and first order coded identification optimization is generated;

Described multilevel coding is applied to one can be keyed in the device.

35, a kind of device of keying in of obtaining of a kind of method as claimed in claim 34.

36, method as claimed in claim 34 is characterized in that: also comprise the following step:

Select the first ergonomics standard of easy touching degree as described multilevel coding;

Select the second ergonomics standard of inquiry rate as described multilevel coding;

The three ergonomics standard of error rate as described multilevel coding searched in selection;

The number of first order coded identification is fixed as 4, so structural precision, the first ergonomics standard of promptly described first order coding is optimization;

The number of second level coded identification is fixed as 16, therefore separates accuracy, the first ergonomics standard of promptly described second level coding is optimization, simultaneously, structural precision, the second ergonomics standard of promptly described second level coding also is an optimization;

Select the second ergonomics standard of English alphabet sequence as described first order coding;

Select the three ergonomics standard of basic English alphabet sequence as described second stage coding;

37, a kind of method that is used for the optimization structure of different coding comprises the following step:

A) select one group of decoding symbol that closely connects each other on statistics, this has comprised following substep:

Select one group of reference data;

Analyze the statistic correlation of symbol with respect to described reference data;

B) select a different cancellation element;

C) number of selection coded identification;

D) select lineup's body engineering standard;

E) arrange these ergonomics standards according to its relation each other;

F) select an optimization method, and

G) use this optimization method.

38, method as claimed in claim 37 is characterized in that: the step of described application optimization method comprises the following step:

1. from one group of candidate code, select a start code;

2. this start code is assigned as a present encoding and a current forced coding;

3. by the paired arrangement of symbol to the appointment of key button, the preferably paired arrangement of all tool possibilities generates one group of new coding by described present encoding being carried out perturbation from described present encoding;

4. detect in the new coding of this group the attribute of each,

5. check a termination criteria, for example restricted standard that further develops of whether having reached;

6. if reached the termination criteria that is proposed, then export this current forced coding,

7., check whether this group newly contains one than the better coding of described current forced coding in the middle of the coding and if do not reach described termination criteria;

8. if there have at least one to be coded in the new coding of this group to be better than described current forced coding, then specify this currently to be encoded to current forced coding,

9., then specify selected at random from the new coding of this group one to be encoded to present encoding if be coded in the new coding of this group goodly without any one than described current forced coding; And

10. repeat these steps in regular turn, from the step of one group of new coding of described generation.

39, method as claimed in claim 37, it is characterized in that: select the step of lineup's body engineering standard to comprise to select one to search error rate and be used as the ergonomics standard and select an inquiry error rate to be used as the step of ergonomics standard, described method also comprises the following step:

1) defines easy touching degree according to the quantized values accepted of searching error rate and inquiry rate;

2) therefore the minimal amount of the required key button of decision is used this minimum key button number and described optimization method and is searched error rate and the acceptable quantized values of inquiry rate;

3) a target typing apparatus be designed to decide under the given situation the maximum number of the key button that can allow; And

4) if the maximum key button number that is determined less than the minimum key button number that is determined, is then carried out the step of described application optimization method.

40, method as claimed in claim 37 is characterized in that: also comprise the step of the different degree coding of described selection reduction, this step further comprises the following step:

1) is chosen in the ergonomics coding as the symbols of representing;

2) method of a coding of selection ergonomics;

3) select the radix of this group coding element, this radix is less than the radix of this symbols;

4) select the reference sequences of a group code thus in the middle of the symbols;

5) data of analyzing this reference sequences are set up the data of the method for coding ergonomics; And

6) be that the coding of the optimal values basically of a method with this coding ergonomics is sought one group of possibility coding.

41, a kind of different degree of application reduction is encoded and is transmitted the method for symbol sebolic addressing, comprises the following step:

1) select one to reduce different degree coding;

2) coding enters one in the code element sequence of selecting to reduce in the different degree coding with this symbol sebolic addressing coding,

3) transmit sequence to a receiver of this code element; And

4), calculate one group of possibility decoding that this is encoded into sub-sequence at receiver end.

42, a kind of device of keying in is characterized in that: include:

One group of exercisable key button that is used to the information of keying in, in this group key button each all contains one group of coupled coding, and will first coding that organize specific in the key button this group coding that interrelates therewith be distributed, then have at least two key buttons to be operated in the middle of this group key button, and to the second or the 3rd coding that organize specific in the key button this group coding that interrelates therewith be configured, have only this specific key button to be operated; And

A different formula is keyed in mechanism, and whether it adopts the second or the 3rd coding when deciding this specific key button in having only this group key button to be operated according to the operation subsequently in this group key button.

43, the device of keying in as claimed in claim 42 is characterized in that: described at least two key buttons comprise at least one auxiliary key button and at least one alphanumeric key button.

44, the device of keying in as claimed in claim 42 is characterized in that: described have at least two key buttons corresponding with an auxiliary key button and an alphanumeric key button, or corresponding with two alphanumeric key buttons.

45, the device of keying in as claimed in claim 42 is characterized in that: each alphanumeric key button has a numeral and at least two letters interrelate with it.

46, the device of keying in as claimed in claim 42 is characterized in that: each alphanumeric key button has a numeral and at least three letters interrelate with it.

Images