HK1140322A - Combined data entry systems - Google Patents

Description

Combined data input system

Related applications:

this application claims priority from the following U.S. provisional patent applications: U.S. provisional patent application 60/816397 filed on 23/6/2006; U.S. provisional patent application 60/819811 filed on 10.7.2006; U.S. provisional patent application 60/837950 filed on 8/15 2006; U.S. provisional patent application 60/843522 filed on 8.9.2006; U.S. provisional patent application 60/856869 filed on 3.11.2006; U.S. provisional patent application 60/859028 filed on 13/11/2006; U.S. provisional patent application 60/874039 filed on 8.2006, 12.2006; U.S. provisional patent application 60/877715 filed on 28.2006, 12.2006; U.S. provisional patent application 60/897807 filed on 26/1/2007; U.S. provisional patent application 60/901132 filed on 13/2/2007; U.S. provisional patent application 60/901978 filed on 16/2/2007; U.S. provisional patent application 60/902383 filed on 20/2/2007; U.S. provisional patent application 60/903571 filed on 26/2/2007; U.S. provisional patent application 60/906178 filed on 9/3/2007; U.S. provisional patent application 60/906491 filed on 12/3/2007; U.S. provisional patent application 60/918656 filed on 16/3/2007; U.S. provisional patent application 60/919340 filed on 21/3/2007; U.S. provisional patent application 60/920376 filed on 26/3/2007; U.S. provisional patent application 60/923733 filed on 16/4/2007; and U.S. provisional patent application 60/927941 filed on 5, 4, 2007, which is hereby incorporated by reference in its entirety.

The present application also relates to prior PCT patent application No. PCT/US00/29647 filed on day 27/10/2000, PCT patent application No. PCT/US02/22385 filed on day 12/7/2001, PCT patent application No. PCT/US04/12082 filed on day 19/4/2004, PCT patent application No. PCT/US05/19582 filed on day 3/6/2005, and PCT patent application No. PCT/US/2006/23524 filed on day 16/6/2006, which are incorporated herein by reference in their entirety.

Technical Field

The present invention relates to data entry. More particularly, the present invention relates to a data input system having features for inputting data using improved predictive and key press/swipe voice interaction methods. The invention also relates to a data entry system that utilizes ambiguous key presses and non-ambiguous key presses/handwritten characters in combination with a predictive character entry system. The present invention also describes different methods of correcting/replacing the words recommended by the data entry system to provide an extremely accurate and simple data/text entry system.

Background

Typical systems and methods for electronically entering characters include the use of a standard keyboard such as a QWERTY keyboard or the like. However, as modern electronic devices become smaller, new methods have been developed for inputting desired characters.

One such method is to use a multi-press system based on a standard telephone number keypad whereby multiple alphanumeric characters are assigned to the same key. One drawback of this system is that it requires multiple presses of a single key in order to enter a particular character, thereby increasing the total number of key presses and slowing the character entry process.

A second approach suitable for entering characters on ever shrinking devices has simply miniaturized the standard QWERTY keypad onto the device. However, such miniaturized keypads are often awkward and do not provide sufficient space between the keys, thus resulting in multiple key presses when only a single key press is desired.

Another method for entering text in a device is a word prediction system. Such systems are typically capable of entering words that are available within the word stock of the system. Entering other words, single letters, numbers, punctuation characters, etc. by such systems can be frustrating and time consuming. In addition, if the system recommends an undesired word, the efficiency of selecting the desired word is reduced because the system may consider all words corresponding to the user's key presses.

Yet another attempt to accommodate character entry on smaller electronic devices is the use of voice recognition software. This method has been used for some time, but has a number of disadvantages. Most notably, voice recognition software does not have the ability to recognize homonyms, and typically requires significant pre-input to the system in order to recognize specific speakers, their particular habits, and speech habits. Moreover, voice recognition software, in an attempt to alleviate these problems, is becoming increasingly large and requires a great deal of processing, and is particularly unsuitable for smaller electronic devices, such as mobile telephones and text pagers, where both power and processing capabilities are limited.

Disclosure of Invention

It is therefore an object of the present invention to provide an improvement over the previously disclosed data entry systems by providing additional enhanced features for predicting and disambiguating ambiguities through the use of various input means including improved keypad, predictive, and voice techniques among other devices. Furthermore, it is an object of the present invention to provide an improved predictive and key press/swipe input system with speech (e.g. voice, lip reading) and/or other additional data input system which facilitates better/faster recognition of words by the predictive data input system of the present invention. It is a further object of this invention to improve the correction capability in such data entry systems.

To this end, in other embodiments, the invention is depicted as a data entry system having a plurality of input receiving devices for entering at least arbitrary characters, wherein each receiving device is assigned more than one character such that a user can enter a desired input using the input receiving device.

Second input receiving means for receiving a second input, wherein the second input is an exact character or characters from the desired input; wherein the system provides a prediction output based on input from at least the plurality of input receiving devices. If the predicted output is different from the desired input and the letter desired to be input (e.g., the first letter) is different from the first letter of the predicted output, the user may input a predefined correction signal to provide the exact (first) character desired to be input, such that the system may provide a second predicted output based on the input previously provided by the user and the correction signal. The invention also describes different simple methods of correcting the words recommended by the system to provide an extremely accurate and simple data/text input system.

Drawings

In the drawings, like components are indicated with the same reference numerals throughout the several views:

FIG. 1 illustrates (a portion of) an exemplary keypad and a portion of a dictionary of the present invention, according to one embodiment of the present invention;

FIG. 2 illustrates four keys of an exemplary keypad of the present invention based on a lip-read version of a data entry system, according to one embodiment of the present invention;

FIG. 3 illustrates one embodiment of assigning multiple letters of a language to the four keys of an exemplary keypad of the present invention;

4-4a illustrate two embodiments of assigning multiple letters of a language to the four keys of an exemplary keypad of the present invention;

FIG. 5 illustrates a system for remapping letters of a key of a keypad on other keys (of the keypad) according to one embodiment of the present invention;

FIG. 5a illustrates a system for on-screen (on-screen) remapping of a plurality of letters of a (hard) key of a keypad on a device, according to an embodiment of the present invention;

FIG. 6 illustrates (a portion of) an exemplary keypad and a portion of a dictionary of the present invention, in accordance with one embodiment of the present invention;

FIG. 7 illustrates a portion of an exemplary keypad and dictionary of the present invention having different mode keys in accordance with one embodiment of the present invention;

FIG. 8 illustrates a portion of an exemplary keypad and dictionary of the present invention, in accordance with one embodiment of the present invention;

FIG. 9 illustrates a portion of an exemplary keypad and dictionary of the present invention, in accordance with one embodiment of the present invention;

FIG. 10 shows an apparatus having an exemplary keypad of the present invention (including ambiguous alphabetic keys) and a QWERTY keyboard with non-ambiguous keys, illustrating the method of the data entry system of the present invention (which may not require speech);

11-11c illustrate a system for remapping letters of a key on a keypad on other keys (on the keypad) according to one embodiment of the present invention;

FIG. 12 shows an exemplary electronic device (e.g., a PC) having an exemplary keypad of the present invention (including ambiguous alphabetic keys) and a split QWERTY keyboard with non-ambiguous keys to illustrate the method of the data entry system of the present invention (which may not require speech);

FIG. 13 shows an exemplary electronic device (e.g., a PDA) having an exemplary keypad of the present invention (including ambiguous alphabetic keys) and a QWERTY keyboard with non-ambiguous keys to illustrate the method of the data entry system of the present invention (which may not require speech);

14-14D illustrate an exemplary electronic device (e.g., a telephone) having a telephone-type keypad with at least some of its plurality of keys replicating the keys of the keypad of the present invention for illustrating the method of the data entry system of the present invention;

fig. 15 shows the keyboard of the invention as shown in fig. 10 used as an external data input unit of a (mobile) electronic device;

FIGS. 16A-16D illustrate an exemplary remapping system of characters of keys on a keypad on other keys (on the keypad) according to one embodiment of the invention;

FIG. 17 shows an exemplary electronic device (e.g., a telephone) having a keypad for entering ambiguous key presses and a telephone-type keypad for entering precise characters of the present invention for illustrating the method of the data entry system of the present invention;

18-18B illustrate a multi-mode electronic device (e.g., a telephone/data-oriented device) including a data input system of the present invention, according to one embodiment of the present invention;

FIG. 19 shows an exemplary electronic device (e.g., a telephone) having a keypad for entering ambiguous key presses and a telephone-type keypad for entering precise characters of the present invention for illustrating the method of the data entry system of the present invention.

Detailed Description

Some previously described features of data entry, such as those in the incorporated references, show data entry systems that utilize a reduced number of keys (e.g., hard keys, soft keys, virtual copy keys, etc.) that have been described heretofore (e.g., a keypad having a reduced number of keys) to enter at least arbitrary characters/text. The data entry system may use a number of symbols comprising at least the letters of the alphabet of at least one language, and wherein the symbols may be distributively assigned to the keys such that at least two of the letters may be assigned to at least one key.

In these systems, a (e.g., predefined) interaction, such as a pressing action on a key, may ambiguously correspond to any of a plurality of symbols/letters assigned to the key (e.g., the key may be referred to as an "ambiguous key"). As previously described, for example, to input one of a plurality of symbols assigned to a key (e.g., such as a letter), a user may provide a predefined interaction with the key, such as a pressing action on the key, and provide voice information corresponding to the symbol (e.g., such as speaking the letter (e.g., the name of the letter)) to select the symbol assigned to the key (e.g., a letter among a plurality of symbols (e.g., a plurality of letters)), and wherein the voice information may be detected and analyzed based on at least one of the user's voice and/or the user's lip motion.

It must be noted that the number of symbols may include substantially all basic symbols, such as alphanumeric characters, punctuation marks, and commands for entering or editing text or manipulating a computer. For example, to enter a punctuation character, a user may press and speak the character on the corresponding key. Other symbols such as macros, partial words, etc. can also be distributively assigned (e.g., a reduced number of keys (e.g., different predefined interactions with the reduced number of keys)).

According to one previously disclosed method, the default of speech, which is detected by the system, can also be considered as speech information, and wherein at least one of the plurality of symbols assigned to a key can be entered at default speech by providing a predefined interaction with the key. For example, in order to input a predetermined punctuation character assigned to a key, the user may press the key corresponding to the character in a predetermined manner without speaking the character.

According to one approach, a reduced number of keys can replicate a telephone-type keypad. The letter assignment may be similar to that of a telephone-type keypad.

According to another approach, the system may use a reduced number of keys (e.g., a keypad), wherein multiple letters of the alphabet may be assigned to at least some of the keys, such as four of the reduced number of keys. The reduced number of keys to be used in the data entry system of the present invention has been described in detail in the patent application incorporated herein by reference.

According to one approach, symbols, such as letters with ambiguous like a predetermined voice, and difficult to distinguish from each other by their corresponding voices, may be assigned to different keys separately from each other or different predefined interactions with the same key.

Also, other related input techniques have been described in detail such as entering at least a portion of a word (e.g., a portion-by-portion data entry system utilizing a press/swipe action for one or more corresponding keys in conjunction with corresponding voice information (e.g., the default predefined nature of the voice can be considered as a voice information)). Also, as mentioned in the prior patent applications, other data entry systems such as full (e.g., hard, soft/on-screen) PC keyboards and/or handwriting recognition systems may be combined/used with the (e.g., press and speak) data entry system of the present invention to enter text/data.

Moreover, as noted in the prior patent applications, derived enhanced data entry systems based on data entry methods have been previously considered and described. For example, the data entry system (e.g., providing precise characters) may incorporate the principles of a word prediction system (e.g., based on pressing actions on ambiguous keys corresponding to characters/letters of a word, and utilizing a word bank and corresponding key press information for each word.) the principles of such word prediction systems are known to those skilled in the art (e.g., T9) to utilize (e.g., very few) few keys (e.g., where it is apparent that at least some of the keys are previously described ambiguous keys) (e.g., and minimize user interactions, such as minimizing user speech or user interactions for selecting a word among candidate words) to provide a very fast and accurate data/text entry system. As previously described, for example, by utilizing the push-and-talk data entry system of the present invention, a user may provide at least one precise character of a word and press a (e.g., ambiguous) key corresponding to the remaining characters of the word to enter the word. By taking into account the at least one exact character of the word, and taking into account key presses corresponding to the remaining characters of the word, the system can predict the corresponding word more accurately.

As previously described, the press/shuffle and voice information (e.g., talk/not talk) data entry system of the present invention may be combined with other data entry systems, such as a word prediction system based on key presses to a keypad (e.g., T9). The principles of a key press based (e.g., key press only) word prediction system are known to those skilled in the art. Also, as previously described, a word of a language may be accurately entered using the press/shuffle and speak data entry system of the present invention by at least one letter (e.g., preferably the starting letter) of a plurality of letters of the word (e.g., or a character (a word may include characters in addition to letters)) and providing key presses (not speaking) corresponding only to other letters (e.g., or characters) of the word for entry. For example, to enter the word "ball," using a telephone-type keypad, a user may enter the initial letter of the word (e.g., "b") by pressing a key corresponding to the initial letter and providing a voice corresponding to the letter, and (e.g., then) pressing keys corresponding to other letters of the word (e.g., in this embodiment, the string "all") without speaking. In this embodiment, providing a sequence of key presses directed (e.g., corresponding, without speaking) only to (e.g., within) all of the letters of the word "ball" (including the letter "b") may also correspond to the word "call". However, since the initial letter "b" is accurately input in the above embodiment, the system does not consider the word "call". According to another embodiment, to enter the word "home" using the same keypad, the user may first press the key corresponding to the starting letter "home" without speaking. He can then press the key corresponding to the next letter "m" and speak that letter (e.g., enter it precisely). The user may then press the key corresponding to the last letter "e" without speaking. The key press sequence may also correspond to the word "gone", but because the system has accurately identified that the third letter of the desired word is the letter "m", the word "home" will be selected/entered accordingly.

As previously described, by providing precisely at least one (e.g., the starting) letter/character of a word (e.g., taking into account a first type of information consisting of at least one letter/character of the word) and providing remaining key presses (e.g., key presses corresponding to the remaining characters of the word) e.g., combining into account a second type of information consisting of the remaining key presses, the system can, in most cases, find a single corresponding (e.g., desired) word within its dictionary that matches the combined information. In this embodiment, by entering the letter "b" precisely, the system is able to select the word "ball" without further user interaction (e.g., such as pressing on the "select" key to scroll through different selections).

It must be noted that, according to one approach, a user may provide a plurality of key presses corresponding to a plurality of letters of a word, and speak any of these letters during the key press even if the user's voice is not synchronized with (e.g., not provided simultaneously with) the corresponding key press.

Also, as previously described, according to one embodiment of the present invention, a user may provide a plurality of key presses corresponding to a plurality of letters of a word and speak the word.

As known to those skilled in the art and as previously described, a word prediction system based on key press information uses at least a dictionary and a corresponding key press database.

According to one embodiment of the present invention, during the input of text, a user may input a word (e.g., a word present in a dictionary, a word outside the dictionary, a character string not present in the dictionary, etc.) by pressing a key corresponding to the letter/character of the word and providing phonetic information corresponding to the word (e.g., the letter of the word). On the other hand, if the user desires, he may enter the words present in the dictionary by entering (e.g., providing) corresponding key presses (e.g., providing key presses corresponding only to letters of a word that may correspond to a large number of words (e.g., selections) within the word stock of the system). For better accuracy and to minimize the number of choices, for example, when the start (e.g., and/or middle, and/or end) key of the word is pressed, the user may speak the start (e.g., and/or middle, and/or end) letter of the word (e.g., providing a key press and speech corresponding to at least one letter of a word, as previously described, the at least one letter may be entered accurately (as described later in this patent application, by the word prediction data entry system of the present invention, other methods may be used during entry of a word to enter one or more accurate characters/letters of the word)). Thus, by considering the at least one exact letter and the key presses corresponding to the remaining characters of the word, the number of choices (e.g., candidate words) can be significantly reduced. (for example, hereinafter, in many paragraphs, such a system that combines the different data entry methods just described (e.g., including predictive methods) is referred to as a "combined" data entry system (of the invention)). Throughout this application, such a system may be referred to as a "combined data entry system", or more precisely, simply as a "(inventive) data entry system". Unless specifically stated otherwise, such terms may be used interchangeably within the context of the present invention.

Having previously described the basic premises of such a combined data entry system in a reference such as that filed by the present inventors and incorporated herein, embodiments of the present invention constitute systems and methods for enhancing such combined data entry systems, as well as correction methods that utilize new words based on information additionally entered by a user through the correction process of the present invention, for example, to replace words provided by the system. The present application also provides additional enhancements, derivations, applications, etc. to the data entry system of the present invention.

The press-and-talk data entry system of the present invention, which permits entry of text by combining at least part of the word data entry method (e.g., referred to as the part-by-part data entry system of the present invention), has been described in detail in prior patent applications filed by the present inventors. According to one embodiment of the word prediction data entry system of the present invention, instead of entering a single character, a user may enter a portion of a word using the press/shuffle and speech data entry system of the present invention for accurately entering at least a (e.g., beginning) portion of the word and provide key presses corresponding only (e.g., without speaking) to other letters of the word. For example, to enter the word "predefined," the user may enter the "pre" portion by a press/swipe action against and speak the corresponding key of the keypad, and then, without speaking, press against the key corresponding to the remaining letter "defined" of the word. It is apparent that during a key press action corresponding to the remaining characters of the word, the user may provide speech (e.g., speak the letters) corresponding to at least some of the characters corresponding to the corresponding key press, if desired, to accurately enter the at least some of the characters.

According to one embodiment of the present invention (although it may reduce the accuracy of the system), in order to enter a precise (e.g., nonsense) character/part of a word/word, a user may speak the part of the character/word without providing (e.g., corresponding) key presses. Thus, to input a word by the word prediction method of the present invention, a user may input at least one precise character by speaking at least one character without pressing a key corresponding to the at least one character, and (e.g., then) press keys corresponding to other characters of the word without speaking. For example, to enter the word "ball," using a telephone-type keypad, a user may enter the initial letter of the word (e.g., "b") by speaking the letter "b" (e.g., without pressing the corresponding key), and (e.g., then) press the key corresponding to the other letters of the word (e.g., in this embodiment, the string "all") without speaking.

Various methods of correcting undesired words recommended by the predictive data entry system of the invention based on user-provided input information (e.g., key press information, and voice information, if provided) have been described previously.

As described in detail, during input of a word, a user may provide first input information (e.g., ambiguous key interactions corresponding to at least some of a plurality of characters of the word) and final non-ambiguous input information (e.g., key presses and speech corresponding to at least one of a plurality of (e.g., other) characters of the word) that includes ambiguous input information.

Embodiment 1 pressing keys 14901, 14902, 14903 without speaking may correspond to a group of words 14908 of table 1 in the system word stock (e.g., reflecting a portion of the word stock), for example, by considering keypad 14900 of fig. 1. The system may recommend the word "was" with the highest priority among the words. This word may not be the word the user wishes to enter.

The word recommended by the system based on the first input information may not be the word the user wishes to input. In order to attempt to replace the word with the desired word, according to one correction method, the user may enter the correction process by providing additional input information (for non-ambiguous entry of the character) such as (e.g., pressing a predefined (e.g., "correct") key and) speaking at least one character/characters of the word (e.g., preferably the starting character, if desired, as described later). By taking into account the first input information and the additional input information, the system may recommend a word from among word groups corresponding to combination information in which the word preferably has a predetermined highest priority among the word groups.

Embodiment 2 in embodiment 1, for example, a voice providing a word such as speaking the letter "y" (during the input of a word (relating to the first input information) or during the correction process (relating to the additional input information)) corresponding to the pressing action provided on the key 14903 may correspond to any one of a plurality of words in the following word group:

“hay”

“may”

“way”

“wry”

the system may predictively recommend the word "may" having the highest priority among the word groups.

The system may also include a navigation system to allow the user to inform the system of navigation within a plurality of words in the word stock corresponding to the input information provided by the user corresponding to the words (e.g., including the first input information, and additional input information, if already provided).

According to a first navigation method, if the word recommended by the system is an undesired word based on the input information provided by the user (e.g., the first input information, and, if provided, the additional input information), the user may provide a predefined command (e.g., a key interaction, a voice command, etc.), such as, for example, providing a predefined press action for a predefined (e.g., "amend") key (e.g., without speaking), and the system may recommend another word in a group of words corresponding to the same said input information provided by the user (e.g., the first input information, and, if provided, the additional input information), where said other word is in the group of wordsA plurality of characters of the character groupWith the next highest priority next to the undesired word.

Embodiment 3 according to this navigation method, in embodiment 1, if the user provides a predefined pressing action with respect to a predefined (e.g., "amend") key, the system may provide a word "has" having the next highest priority next to the word "was" among a plurality of words of the word group (e.g., the word group corresponding to the first input information provided by the user).

Embodiment 4 according to this navigation method, in embodiment 2, if the user provides a predefined pressing action with respect to a predefined (e.g., "amend") key, the system may provide a word "way" having the next highest priority next to the word "wa" among a plurality of words of a corresponding word group (e.g., a word group corresponding to the first input information and the additional input information provided by the user).

According to one embodiment of the invention, if the input information (e.g. provided by the user) corresponding to the first character of the desired word is ambiguous, according to the second navigation method, if the system is based on the input information provided by the user(e.g., first input information, and additional input information, if provided)) the recommended word is an undesired word, the user may provide a predefined command (e.g., a key interaction, a voice command, etc.), such as, for example, providing a predefined press action for a predefined (e.g., "amend") key (e.g., without speaking), and the system may recommend another word in the set of words corresponding to the same input information (e.g., first input information, and additional input information, if provided) provided by the user, wherein the other word begins with the same starting character of the undesired word, and wherein the other word is preferably at the same starting character of the undesired wordA plurality of characters of the character groupWith the next highest priority next to the undesired word.

Embodiment 5 according to this navigation method, in embodiment 1, if the user provides a predefined pressing action for a predefined key, the system may provide a word "way" having the next highest priority to the word "way" only among a word group corresponding to the first input information provided by the user (e.g., a key press corresponding to a character of the word) and wherein the word starts with a letter "w" (e.g., in this embodiment, the word group includes the words "way", "wry", "wag).

Embodiment 6 according to this navigation method, in embodiment 2, the word "way" is a unique word corresponding to the first input information provided by the user (e.g., a key press corresponding to a character of the user-provided word) and the second input information (e.g., uttering the third letter "y"), and wherein the word starts with the letter "m" (e.g., the first letter of the system-recommended word). In this embodiment, pressing the navigation command may be useless.

The navigation command may be repeated several times when needed (e.g. pressing the command/correction key may be repeated several times), wherein each time the system may provide another word in the corresponding word group, wherein the other word preferably has the highest priority next to the last recommended word.

Example (b):

in embodiment 3, the system may recommend the word "may" by providing an additional predefined press action against a predefined (e.g., "amend") key. Nonetheless, more than one additional predefined press action for a predefined (e.g., "modify") key may still correspond to the word "may". And so on.

In embodiment 4, the system may recommend the word "hay" by providing an additional predefined pressing action for a predefined (e.g., "amend") key. Nonetheless, more than one additional predefined press action for a predefined (e.g., "modify") key may also correspond to the word "wry" (e.g., in this embodiment, the word "wry" is the last word in the corresponding group).

In embodiment 5, the system may recommend the word "wry" by providing an additional predefined pressing action for a predefined (e.g., "amend") key. Nonetheless, by providing more than one additional predefined press action for a predefined (e.g., "amend") key, the system may also recommend the word "wag" (e.g., in this embodiment, the word "wag" is the last word in the corresponding group).

According to one embodiment of the invention, the additional input information (e.g. speech) may be provided at the end of the first input information while the cursor is positioned at a predetermined location such as just after the last character of the corresponding word (e.g. according to one approach, to avoid the system erroneously perceiving/considering the undesired speech, the desired speech may be provided (e.g. started to be provided) within a predetermined time delay (lapse soft time) after positioning the cursor at said location). If the user speaks after the delay, the system will not take the speech into account. If the user wishes to provide additional speech information, he can move the cursor to another location (e.g. by pressing the space bar) and then return it again just after the last character of the word (e.g. by pressing the "Bk Sp" key and (starting) speaking within the time delay).

According to one embodiment of the invention, a navigation process (e.g., a pressing action on the "correction" key) may preferably be provided/taken into account by the system when positioning the cursor just after the last character of the corresponding word.

According to one embodiment of the invention, additional (input) information (e.g., speech) may be provided by providing a predefined (e.g., type) command (e.g., a tap, a smooth sound command, etc.), such as providing a predefined type of press action for a predefined (e.g., "amend") key and providing the additional information (e.g., substantially simultaneously). Providing the same/another predefined command without providing speech, such as providing the same/another predetermined type of pressing action for the same/another predefined (e.g., "amending") key, may correspond to the navigation process within the corresponding word. For example, pressing keys 14904, 14902, 14903 in keypad 14900 of fig. 1 may correspond to words 14909 of table 2 (e.g., reflecting another portion of the word stock of the system). The system may recommend the word "its" with the highest priority among the plurality of words (e.g., based on frequency of use). If the word is an undesired word while the cursor is pointing at the word (e.g., while, for example, the cursor is positioned just after the last character of the word, or by positioning the cursor just after the last character of the word), the user may press a predefined (e.g., "fix up") key and preferably while speaking/pronouncing the letter "d". By considering the pressing action on the predefined (e.g., "amend") key and detecting the user's voice, the system may understand that the user wishes to provide additional information corresponding to the word. By analyzing the speech, the system may recommend a word having the highest priority among the words in the table (e.g., corresponding to the key press) among the words beginning with the letter "d". The system may recommend the word "day". If the word is not the desired word, the user may press the predefined (e.g., "amend") key (e.g., once) without speaking. By considering a press action for a predefined (e.g., "fix") key and not detecting speech, the system understands that the user may wish to navigate to a next word in the table (e.g., corresponding to the key press) within a plurality of words beginning with the letter "d", and wherein the next word has a next highest priority among the plurality of words next to the word "day". In this embodiment, the word is the word "dry". The system may recommend the word to the user. The user may accept the word by providing another (e.g., different) action, such as, for example, providing a space character or proceeding with the entry of the next word.

As previously mentioned, in any of the embodiments or methods (such as the embodiments described above) relating to the (e.g. predictive) data input system of the present invention, instead of/in addition to providing speech information for inputting non-ambiguous characters (e.g. ambiguous characters of a word), other input methods such as handwriting input and recognition systems, full/pc keyboards, etc. may be used for inputting non-ambiguous characters. It has to be noted that the predefined interaction with a key may be a predefined action such as a single press action for said key, or a double press action for said key, or a swipe action for said key, or a long press action for said key, or a press-and-hold and press of another key. These events have been described in prior patent applications previously filed by the present inventors.

As previously described, after inputting first input information (e.g., key presses, and voice (if desired/required)) corresponding to a word using the predictive data entry system of the present invention, if the recommended word is not the word that the user wishes to input, the user may enter the correction process by providing additional information, such as at least one of a plurality of letters/characters (the starting letter/character) that uttered the desired word.

According to an embodiment of the invention, the additional information (e.g. speech) may be provided when pointing/selecting the word, such as for example when positioning a cursor at a predetermined position related to the undesired word (such as for example just after the end character of the undesired word, or on the undesired word, or after a space character after the word, etc.) and providing a predetermined interaction with the (e.g. predetermined) key. For inputting the additional (e.g. speech) information several key interaction methods may be considered in combination with the additional information

According to a first approach, a user may provide a single predefined key interaction, such as a single-press action against a predefined (e.g., correction) key, and provide speech corresponding to a character (e.g., preferably the starting character) of a desired word. To enter additional information corresponding to each of the other (e.g., following) characters of the desired word, the user may thus repeat the predefined key interaction and speak the corresponding character. For example, by considering the keypad 14900 of fig. 1 and by considering the word stock 14909 of the system, to enter the word "dry", if the user enters the first input information by pressing the corresponding key without speaking, the system may recommend the word "its" that the user does not wish to enter. Next (e.g., when pointing to the word, or when the cursor is positioned just after the word, or on the word, etc.), the user may, for example, provide two predefined press actions (e.g., two single press actions) for the predefined (e.g., correction) key and provide the second information by speaking the letters "d" and "r" at (e.g., preferably substantially the same time). The system may understand that the user wishes to replace the word "its" with another word corresponding to said first input information (e.g. the same corresponding key press initially provided by the user) and that the speech of its two starting characters corresponds/is the speech provided by the user. The system recommends the word "dry" as a unique word in the word stock corresponding to the combined first and second input information provided by the user.

According to a second method, the user can provide a press-and-hold action on the correction key and speak as many characters as he wishes, and then release the key.

According to a third method, the user may provide a single predefined key interaction and speak the desired word.

As previously described, speech may also be detected and recognized based on the lip reading system. Also, as previously mentioned, letters having ambiguous similar voices associated with each other, such as a language, may be assigned to different input signals (such as to different keys) separately from each other for better accuracy. Fig. 2 illustrates an embodiment of a distribution of english letters over four keys 15001-15004 of (e.g., partial) keypad 15000 based on these principles. For example, the letters "B", "M", "P", "W" (a predetermined period in the process of speaking each of them requires the user's lips/mouth to be in a closed position) may be assigned to different keys. Also, as an embodiment, the letters "T", "D", "N", "L" (between the user's teeth where the user's tongue is required to be in a substantially closed position for a predetermined period of time in saying each of them) may be assigned to different keys. Similarly, other letters with similar lip reading characteristics have been assigned to different keys separately from each other. A dictionary/word bank of a key press based word prediction data entry system typically contains only about 25000 words. The predictive data entry system of the present invention permits the system to consider only a few words corresponding to input information provided by the user. Thus, a larger dictionary may be considered where the system may still have a high level of performance.

According to one embodiment of the invention, the dictionary database of the system may contain words of more than one language, profession, environment, etc.

According to another approach, multiple separate dictionaries (e.g., dictionaries of different languages, dictionaries of different specialties, dictionaries of different environments, etc.) may be used with the predictive data entry system of the present invention, wherein a user may switch within the word stock (e.g., even within the same document) using a switching means such as providing predefined interactions with predefined keys or voice commands.

As previously described, during entry of a word using the combined (e.g., including word prediction) data entry system of the present invention, a user may enter a letter/character by pressing a key corresponding to at least one character, such as a letter, and providing speech corresponding to the letter/character (e.g., speaking the letter/character). According to an embodiment of the present invention, the key depression and the voice may preferably be provided simultaneously.

It may happen (e.g., when the user is typing quickly) that a user voice corresponding to one of the consecutive key presses is provided during more than one of the (e.g., quick) consecutive press actions (e.g., the voice may cover more than one of the consecutive key presses for the time of the press action). This can lead to ambiguity for the system, which may erroneously believe that the user's voice corresponds to the continuous key presses of the voice overlay.

According to one embodiment of the invention, to help the system better understand the user's intent:

if a user provides speech during a fast consecutive key press (e.g. the time delay interval between each of two of the key presses is smaller than a predetermined time delay), wherein the speech does not cover the part of the fast consecutive key presses provided after the speech (when the speech is started), it is predefined that the system may consider that the speech of the user may correspond to only one of the consecutive key presses, wherein preferably the key press is predefined to be a key press during which the speech of the user is to be started to be provided.

-if the user wishes to provide speech corresponding to two or more quick consecutive key presses, the user may:

-providing at least a predefined time delay interval/pause when ending the input of a fast succession of key presses corresponding to said speech, or:

-pressing and speaking at a moderate speed by providing at least a predetermined delay/pause between each of the successive key presses (e.g. and each of their corresponding voices).

The above rules may also be applied to enter a portion of a word and/or to enter a mixture of characters and a portion of a word. For example:

if the user provides a speech corresponding to a character string, such as a part of a word, during a fast consecutive key press, wherein the speech does not cover the part of the fast consecutive key press provided after the speech (when the speech is started), it is predefined that the system may consider that the user's speech may correspond to only one of the consecutive key presses (e.g. the system may suggest that the speech is predefined assigned/predefined to be represented by the key press). Wherein the preferably predefined key press is a key press that starts a period during which the user's speech is to be provided.

If the user wishes to provide speech corresponding to two or more quick consecutive key presses (e.g. the speech may be such as speaking at least two characters, or two parts of at least one word, or at least one character and a part of a word, etc.), the user may:

-providing at least a predetermined time delay interval/pause at the end of the input of a fast succession of key presses corresponding to said speech, or;

medium speed presses and speaks by providing at least a predetermined delay/pause between each of the consecutive key presses (e.g., and each of their corresponding voices).

During the provision of several rapid successive key presses, the user may provide more than one voice (corresponding to more than one of the key presses). It is obvious that the above principle automatically responds/applies to the situation.

(e.g., if the portion of the consecutive key presses provided during and between every two consecutive voices are not covered by those voices (e.g., either of the two), the system may consider, based on the principles described above:

-a first voice of the two user voices may correspond to only one of the consecutive key presses, wherein the key press is preferably predefined as a key press during which the first user voice is to be provided to the beginning, and;

-a second speech of said two user voices may correspond to only one of said consecutive key presses, wherein said key press is preferably predefined as a key press during which said second user voice is to be provided at the beginning).

It should be appreciated that any of the user's voices may be speaking a corresponding character, such as a letter, or speaking a corresponding portion of a word, etc.

As an example of the above principles, by considering the (local) keypad 15000 of fig. 2, to enter the phrase "we are meeting", if the user presses only the keys corresponding to the characters of the phrase, the system may erroneously output the phrase "he are meeting" (e.g., the word "he" has a predefined higher priority than the word "we"). It is obvious that, after a few practices, the user knows in advance that, in order to enter the word "we", according to one method of the invention, while pressing the key corresponding to the letter "w", he can speak said letter and, without speaking, press the key corresponding to the next letter "e". It is obvious that after a considerable amount of practice, the user is also able to type quickly. During the entry of the phrase, the user may quickly tap the keys corresponding to the characters of the phrase. The user may speak the letter "w" while pressing the key corresponding to the letter. Because the user types quickly, his voice corresponding to the letter "w" may also override one or more subsequent ones of the consecutive key presses in addition to the corresponding key press 15001. With the above-described embodiments, by recognizing the speech provided during a fast succession of key presses, the system can understand/decide that the speech belongs to the key press 15001 during which speech is to be provided at the beginning. This permits the user to type quickly.

The above-described embodiments/methods may significantly improve the speed of the combined (e.g., including word prediction) data entry and correction system of the present invention. This is because during the input of text, most words of the text (e.g., present in the word stock of the system) are typically output by the system based only on the corresponding key presses provided by the user. Additionally, with/in mind the revision process of the system, generally, a user rarely needs to speak more than one character of a word (e.g., as present in the word stock of the system) to enter the word. In addition, when a user speaks a sequence (e.g., two or more) of consecutive/successive characters, typically, he naturally speaks them slowly and typically naturally provides a pause when ending the consecutive/successive speech. By combining the methods just described in the combined (e.g., including word prediction) data entry and correction system of the present invention, a very fast, simple and very natural mobile data entry and correction system can be provided. Such a data entry system is even easier than with a standard (e.g., QWERTY) keyboard, so after using it during a small period of time, it can also become the preferred data entry system for a stationary device such as a desktop computer.

It must be noted again that when modifying/changing the undesired words recommended by the system (e.g., corresponding to the initial input information, which may include key presses, and the speech (if provided)), the user may provide additional input information, such as speech corresponding to one or more characters/letters of the words. For example, to enter speech for more than one character of a desired word, a user may press a predefined interaction with a predefined (e.g., "fix") key a number of times (e.g., "n") corresponding to the number of multiple (e.g., "n") (e.g., preferably consecutive starts) characters/letters of the desired word and provide speech corresponding to the characters/letters (e.g., by speaking the characters/letters).

If the system uses the same correction key and key interaction for a first correction process consisting of a process that provides multiple (e.g., "correct") key presses-and-voices corresponding to multiple (e.g., preferably at least the initial) letters/characters of the desired word, and a second correction process consisting of a process that provides multiple key presses-and-voices, wherein each of the multiple key presses-and-voices is intended to inform the system of words in the recommended word group that correspond to the initial/first (e.g., key press) information (e.g., corresponding to the characters of the desired word) and at least additional voice information provided during the ("correct") key press (e.g., the word having the highest priority among the words), then to help the system distinguish between the two correction processes, the following rules may be applied:

-providing the predefined key interaction and speech quickly (e.g. by providing a delay/pause between successive correction key interactions, wherein the delay/pause is shorter than a predefined delay) may correspond to a first correction procedure (e.g. a user may provide speech corresponding to all characters/letters of a word even while providing corresponding interactions with the ("correction") key as many times as needed (such as the number of times the key is pressed corresponding to the number of characters of the desired word), if desired;

providing said predefined key interactions and speech at a medium speed (e.g. by providing at least said predetermined delay between successive modifier key interactions) may correspond to a second modification procedure.

Each additional press action directed to a ("correction") key without speaking (e.g., in this case interpreted as a "next word" key) may inform the system of navigation within the plurality of words corresponding to the initial key press of the character corresponding to the desired word, and additional speech information provided by the user at least during the correction process corresponds to the word, so that the system may recommend the word having the next highest priority among the plurality of words.

According to one embodiment of the invention, to input speech during a correction process for correcting/replacing a recommended word (e.g., to help the system identify at least some of a plurality of precise characters of a desired word), a user may speak a portion/word of the character/word without providing any key interaction (e.g., the system may not have a predefined (e.g., "correct") key, or the user may not use a predefined (e.g., "correct") key). To this end, the user may select/point to an undesired recommended word in a predetermined manner and then provide the voice. The user may use predefined ("modifier") keys for navigating within a plurality of words corresponding to initial input information (e.g., key presses, and final speech (if provided)) for the characters (initially provided by the user) corresponding to the words, as well as speech information provided during the modification process.

It has to be noted that according to an embodiment of the present invention, during the correction process of the present invention, a first predefined key (e.g. or interacting with a first predefined key of the predefined keys) may be assigned to the correction process for entering additional speech information of the user corresponding to the characters of the desired word, while a second predefined key (e.g. or interacting with a second predefined key of the predefined keys) is assigned to the correction process for navigating (e.g. serving as "next word key") within a plurality of words corresponding to the initial input information provided by the user and, if provided, the additional input (e.g. speech) information provided during the correction process.

According to another embodiment, a predefined means, such as a predefined key or voice command, may be used for switching between the first and second correction process.

According to another embodiment, when the same revision key is used (e.g., or the same predefined interaction with a revision key) for each of the revision processes, a predefined means such as a predefined key or voice command may be used for switching between the first revision process and the second revision process. For example, a first correction process may use (or interact with) predefined keys during the input of additional (voice) input information, while a second correction process may use (or interact with) the same predefined keys to conduct a navigation process through a plurality of words corresponding to the initial input information and additional input (e.g., voice) information (provided during the correction process) provided by the user.

It must be noted that although the term "recommended word" has been used in many paragraphs of the patent application filed by the present inventors, the term is intended to describe a word that is or is to be recommended by the predictive data entry system of the present invention. For example, sometimes during the input of ambiguous and non-ambiguous input information corresponding to a word, the user may know in advance which word the system will select. The user may not want the word and may enter the correction process immediately after entering the input information (e.g., before the word is recommended to the user). Also, for example, according to another embodiment, the system may not provide any (e.g., recommended) words until the user provides a predefined command, or the like.

It must also be noted that all key interactions during the modification process of the present invention may be replaced by or combined with other data input methods such as voice commands. For example, in order to navigate through a number of words corresponding to the initial input information and additional (speech information), if already provided, the user may use a predefined voice command, such as for example say "next" (e.g. after pointing to the word). Also, for example, to enter additional (e.g., speech) input information corresponding to one or more characters of the word, a user may speak the characters, a portion of a word, etc. (e.g., after pointing to the word).

It must also be noted that the initial (e.g., first) input information provided by the user includes key presses corresponding to at least a portion of the plurality of characters (e.g., all characters) of the word, and ultimately includes speech corresponding to a portion of the plurality of characters of the word (e.g., if provided). According to an embodiment of the present invention, during the correction process of the present invention, it is predefined that the system does not consider a part of the speech included in the initial information corresponding to a part of the plurality of characters of the word (for example, the speech provided mainly for the middle character).

According to one embodiment of the invention, during the correction process, when the user navigates through a plurality of words corresponding to the initial input information and, if provided, additional (e.g., speech) information provided, if already provided, during the correction process, after the system recommends the last word (e.g., a word having the lowest priority among the plurality of words corresponding to the initial and additional (e.g., if provided) input information), an additional predefined command (e.g., an additional predefined interaction such as a predefined interaction with a predefined (e.g., "correct"/"next") key or a voice command) may cause the system to (e.g., periodically/cyclically) recommend the word having the highest priority among the group of words corresponding to the information. Each additional identical predefined interaction may inform the system to recommend the word with the next highest priority among the remaining words of the word group.

It has to be noted that, in the above-described embodiment, according to one method, after the system recommends the last word of the word group (e.g., the word having the lowest priority among the plurality of words corresponding to the initial and additional (e.g., if provided) input information), the user may provide voices (e.g., speak them one by one) corresponding to all characters of the desired word by providing an additional predefined command (e.g., an additional predefined interaction such as a predefined interaction with a predefined (e.g., "correct"/"next") key or a voice command).

According to one embodiment of the invention, during the correction process, when the user navigates (presses the "next"/"correction" key without speaking) through a plurality of words corresponding to the initial input information, and (if provided) additional (e.g., speech) information provided during the correction process (if provided), appending a predefined command (e.g., appending a predefined interaction such as a predefined interaction with a predefined (e.g., "correct"/"next") key or a voice command) after the system recommends the last word (e.g., a word having the lowest priority among the plurality of words corresponding to the initial and additional (e.g., if provided) input information) may cause the system to recommend only the word having the highest priority among the group of words corresponding to the initial key press information. Each additional identical predefined interaction may inform the system to recommend the word with the next highest priority among the remaining words of the word group.

As mentioned in the different patent applications filed by the present inventors, input information provided by a user via an electronic device may be (remotely) transmitted to a central computing device, such as a server computer, in which at least a part of the system, such as a data processing/identification module in the system, corresponding to the input information provided by the user is installed. After processing the user's input, the results (e.g., output) may be sent to a destination such as the user's electronic device (e.g., for the user to verify accuracy) or another destination. For example, if the user's electronic device is a cellular telephone with limited processing power and/or memory size, the user may enter key presses and speech (corresponding to the desired text) in accordance with the principles of the data input system of the present invention. He may then send the information to a remote computer via the cellular network or another network. After receiving the results/output text and verifying its accuracy, the user may send the text to his desired destination, or he may input a confirmation signal to a remote computer to cause the computer to send the text results to the user's desired destination, etc.

The different methods of distributing/assigning english alphabets to the four keys of the keypad have been illustrated and described in detail previously. FIG. 3 illustrates another embodiment of letter distribution/assignment to four keys based on principles such as at least some of the following principles:

-assigning letters with ambiguous similar speech to different keys

The letters are assigned in such a way that most words in a text do not require speech input information (corresponding to a letter/part of a word)

To assign letters to a plurality of keys, and the keys are positioned in relation to each other so that they can be used widely (as much as possible) uniformly.

In this embodiment, the distribution/assignment of these letters is similar to those of the keypad of FIG. 1, with slight modifications. The letter "V" is moved to the key 15103 (e.g., the letter "V" and the letter "D" may be misrecognized in relation to each other when using a low end/far away microphone or when whispering, etc.), and the letter "L" is moved to the key 15102 (e.g., the letter "L" and the letter "O" may be misrecognized in relation to each other when using a low end/far away microphone or when whispering, etc.).

As previously mentioned, if the user also provides speech (e.g. pronunciation) corresponding to (e.g. to be considered in combination with) at least one of the press actions while providing the quick press actions, the speech may cover a part of the quick press actions after those press actions that are desired to be covered. Different approaches to solving this problem have/may be proposed. For example, according to one approach, the system may consider/combine at least one or more of the pressing actions covered by the speech:

one (e.g. start)

N-1 (e.g. Start)

-N (all)

Other quantities

And the speech and decide which of them provides the best result. Also, as an embodiment, according to another method for solving this problem, it has been described in a previous provisional patent application filed by the present inventors. It is apparent that the system can consider/combine one or more of the methods at the same time to generate the best results.

As described in detail in various patent applications filed by the present inventor, the data entry system of the present invention discloses a system for entering at least any character and/or a Portion of one or more words (e.g., corresponding to a syllable) using a reduced number of keys (e.g., hard keys, soft keys, virtual copy keys, etc.) (e.g., a keypad having a reduced number of keys) (i.e., a Portion of a word may be written as a "Portion of a word" thereafter). Moreover, as described above, the data entry system of the present invention may use a number of base symbols in at least one language to enter or edit text or to manipulate a computer. The number of symbols may comprise substantially all of the base symbols, such as at least one of the following groups:

text symbols such as letters, parts of words (e.g. corresponding to syllables of words),

-an Arabic number, wherein,

-a punctuation mark,

-commands, functions, etc.

The symbols can be distributively assigned to the plurality of keys in such a way that at least two of the text symbols are assigned to at least one key. As previously described, to accurately input one of a plurality of (text) symbols (e.g., letters, a portion of a word) assigned to a key, a user may provide a predefined interaction with the key, such as a press action on the key, and provide speech information corresponding to the symbol (e.g., such as speaking a letter (e.g., a letter name), speaking a portion of a word (a syllable)), to select a symbol (e.g., a letter) among the plurality of symbols (letters) assigned to the key (the predefined interaction with the key), and wherein the speech information may be detected and analyzed based on at least one of a voice of the user and/or a motion of lips of the user. The precise text symbol may be a portion of a word entered by the combined (e.g., including predictive) data entry system of the present invention, wherein a user may not provide speech corresponding to some of a plurality of key presses corresponding to some of a plurality of letters/portions of the word. Data entry systems and enhanced systems therefor have been described in this application and various other patent applications filed by the present inventors. For example, according to one method, to enter a word or a portion of a word, a user may press at least one of a plurality of corresponding keys (e.g., start) of the portion and speak the portion. It is apparent that the combined (e.g., including predictive) data entry and modification system of the present invention may be used to enter text in languages based on different letters (e.g., based on different alphabets), such as english, french, korean, etc., as described above.

Moreover, it is apparent that the combined (e.g., including predictive) data entry and correction system of the present invention may also be used to enter ideographic text such as Chinese, Japanese, etc. The only difference may be that each word/entry of the morphological database (e.g., a string of text symbols (e.g., letters of a roman-based language) may in this case be a phonetic symbol corresponding to a morphological character/word as described later in this patent application) represent a morphological character/word, whereby the character/dictionary of the morphological language may include the corresponding (e.g., written) morphological character/word, wherein after input information (e.g., key presses, and speech (if provided)) is entered, instead of printing the predicted/selected character/word in the phonetic symbol, the system may print the corresponding morphological character/word.

Each written character in a morphological language such as chinese may be represented by a phonetic representation of the character. These are several phonetic systems for inputting chinese characters such as pinyin and bopomofo.

Pinyin is the latin speech system of chinese for representing chinese speech using european letters. Of all 26 letters of the english alphabet, 25 are used for pinyin. The pronunciation "V" is not a phoneme used in chinese. Although the letters are the same, the sounds they represent are different in some cases. For this reason, there is virtually no easy way to accurately determine how the chinese characters are pronounced based on the pinyin, but it is very easy for english speakers to guess.

In addition to replacing the roman letters of the english alphabet, the chinese pinyin letters are another standard for representing chinese pronunciations that uses chinese derived letters for the alphabet. Because the chinese pinyin characters have 37 normal characters, a more complete chinese spelling is provided than 26 english letters.

In Chinese, syllables are composed of consonants, vowels, andtone(s)And (4) forming. According to the traditional notation (traditional notation), each Chinese syllable consists of a "beginning" and a "ending".

English consists of vowels and consonants. In Chinese, each syllable consists of an initial and a final.

The initial is substantially equivalent to the english consonant. The first 21 symbols of the Chinese phonetic alphabet

The tail is substantially equivalent to the english vowel. A tail may be combined with other tails to form a composite tail. Although the Chinese Pinyin alphabet contains 16 Pinyin coda, Pinyin also includes additional compound coda. There are also several new pronunciation rules on how to combine the initial and final sounds together.

Each character in chinese may be represented by a combination of one, two, or more phonetic symbols.

To input a morphological character such as a chinese character represented by a number of phonetic symbols, the corresponding key may be pressed (respectively). Because more than one phonetic symbol may be assigned to a key, the key that is pressed may correspond to more than one corresponding character. This can lead to ambiguity. This ambiguity can be resolved by using the principles of the data entry system of the present invention.

To input text in a formal language by the combined data input system of the present invention, letters (e.g., phonetic symbols) may be distributively assigned to the keys of the keypad based on the symbol assignment principles of the data input system of the present invention described in different (e.g., this and other) patent applications filed by the present inventors. For example, phonetic symbols with ambiguous similar speech may be assigned to different keys separately from each other or to different interactions with the keys. Also, for example, the phonetic symbols may preferably be separated from each other in different groups based on the principles of the data input system of the present invention, so that as few keys as possible are used, such as preferably four keys (or at least a part of a plurality of keys using a telephone type keypad).

The predictive morphological data entry system of the present invention may have words in which each morphological word in the database of morphological languages may have its corresponding phonetic symbol information in addition to corresponding key press (e.g., value) information. The system may also include a (e.g., single) character database and their corresponding key presses to enter/select (e.g., single) characters (e.g., characters of a word) based on the user's input information (e.g., key presses, and corresponding speech (if desired)). The system may also include a (e.g., single) syllable-based stem database and their corresponding key presses to input/select (e.g., a single) syllable-based stem (e.g., a stem of a word) based on the user's input information (e.g., key presses, and corresponding speech, if desired).

It is apparent that all other data entry, editing, correction and enhancement described in this and other patent applications filed by the present inventors can be applied to this level of data entry (e.g., in appropriate paragraphs/embodiments of the patent application, the terms "letter" and/or "character" can be considered/relate to phonetic symbols). For example, to input a phonetic symbol, the user may press the corresponding key and provide the corresponding predefined voice. Also, as an embodiment, when inputting with the prediction data of the present invention, if the recommended word is not the desired word, the user may enter the correction process by (e.g., pressing the correction key) and speak the initial phonetic symbol of the word.

For better accuracy, the system may also use corresponding linguistic rules, such as the combination of the initial and final notes described above.

According to a data input method, for example, in order to input an ideographic character, while a user presses a key corresponding to a phonetic symbol, the user may provide the voice of/corresponding to at least one of the phonetic symbols for better accuracy. If desired, the user may provide an end signal, such as pressing a predefined key, upon ending entry of input information (e.g., key presses, and speech if provided) corresponding to the ideographic character. The system compares the input information provided by the user with (stored) information of ideographic characters of the dictionary to recommend a corresponding character. If the recommended character is not the character that the user wishes to enter, the user enters a correction process so that the system can recommend another character. The process of recommending ideographic characters to a user and correcting/modifying for navigation within a plurality of corresponding characters of a corresponding dictionary may be similar to the process described (e.g., for word input) in the present application and prior patent applications filed by the present inventors. With the present method, a user can (e.g., precisely) enter ideograms character-by-character.

According to a data input method, for example, in order to input an ideographic word, while the user presses a key corresponding to a phonetic symbol of the word, the user may provide the voice of/corresponding to at least one of the phonetic symbols for better accuracy. If desired, the user may provide an end signal, such as pressing a predefined key, upon ending entry of input information (e.g., key presses, and speech if provided) corresponding to the ideogram. The system compares the input information provided by the user with (stored) information of ideographic words of the dictionary to recommend corresponding words. If the recommended word is not the word the user wishes to enter, the user may enter a correction process so that the system may recommend another character. The recommendation of an ideographic word to a user and the correction/modification process for navigating within a plurality of corresponding words of a corresponding dictionary may be similar to the processes described in the present application and prior patent applications filed by the present inventors (e.g., described for word input).

According to another data entry method, for example, to enter ideographic characters/words, the user may provide the voice of a syllable (e.g., the phonetic symbol corresponding to the syllable) for better accuracy while the user presses the key corresponding to the phonetic symbol. If desired, the user may provide an end signal, such as pressing a predefined key, upon ending the input of input information corresponding to the syllable (e.g., key presses, and speech if provided). The system compares the input information provided by the user with the information of the stored information of the syllable dictionary to recommend/select the corresponding syllable.

Different methods of distributing/assigning english letters to the four keys of the keypad have been illustrated and described in detail previously. Fig. 4 illustrates another embodiment of distributing/assigning letters to four keys based on principles such as at least some of the following principles:

-assigning letters with ambiguous similar speech to different keys

The letters are assigned to a plurality of keys in such a way that most/many words of a text can be input based on key interactions (e.g. key presses) only (e.g. when the corresponding key is pressed, speech input information corresponding to the letter/part of the word is not needed in most cases)

Assigning letters to a plurality of keys, and positioning the keys in relation to each other such that (e.g. during entering text) most of them can be used (e.g. pressed) evenly.

In the (e.g., partial) keypad 152000 of fig. 4, the distribution/assignment of those letters is similar to the distribution/assignment of the (e.g., partial) keypad of fig. 3, which is slightly modified. The letter "M" is moved to the key 152002, and the letter "L" is moved to the key 152001 (e.g., swapping them). In this arrangement, the keypad is largely responsive/corresponding to the principles described above.

In the (e.g., partial) keypad 152110 of fig. 4a, the distribution/assignment of those letters is similar to the distribution/assignment of the slightly modified (e.g., partial) keypad of fig. 4. The letter "M" is moved to the key 152113. This keypad is also responsive to the above principles.

It should be understood that other arrangements for assigning any small number of keys with letters in any language based on any of the above principles (e.g., preferably all) may be considered by those skilled in the art. For example, other slight modifications derived from the alphabetic arrangement on the four keys shown here and in the prior patent applications filed by the present inventor may be considered.

Different methods of entering precise characters without providing speech have been described previously. According to one embodiment of the present invention, the keypad of the present invention may be designed such that when a user presses (e.g., and preferably holds) a first key corresponding to/representing a desired character (e.g., character) he presses a second corresponding (e.g., predefined) key to accurately enter the character. As an embodiment, in addition to a (e.g., short) single press (e.g., for entering an ambiguous letter) on a (first) key of the (e.g., ambiguous) keypad of the present invention, another predefined interaction with the key, such as a long press, press-and-hold, etc., may also inform the system that the user wishes to enter an exact character (e.g., letter) without speaking. The user may (preferably simultaneously) press the second corresponding key to enter the desired character. Another pressing action on the (e.g., first) key corresponding to the (e.g., ambiguous) character/letter may (predesignatedly) map each of at least some of the plurality of (ambiguous) characters/letters of the key to a different key of the keypad. The user may then press a respective different (second) key corresponding to said character/letter (preferably simultaneously).

Fig. 5 illustrates an embodiment of the keypad of the present invention in which letters are assigned to four of its keys based on the principles of the data entry system of the present invention as previously described. The letters assigned to each of the keys 15201-. As an example, a single pressing action on any of the keys may provide an ambiguous character of a word. Another (predefined) type of interaction for each of the keys 15201-15204 may map the letters on that key to a different key among the keys 15201-15208. As an example, the mapped letters for each of the keys 15201-15204 are shown on the lower regions of the keys 15201-15208. For example, to enter the letter "L," the user may press key 15201 and (preferably simultaneously) press key 15204. Also, as an embodiment, to enter the letter "B", the user may press key 15202 and (preferably simultaneously) press key 15208. Mapping characters/letters corresponding to each set of (ambiguous) characters/letters of a key may be printed on corresponding mapping keys of the keypad and/or they may be dynamically printed on corresponding mapping keys of the (virtual) keypad each time a corresponding first key is pressed. For easier interfacing/interaction, each set of ambiguous characters of a key and its corresponding mapped character on the corresponding key may be printed in a different color, font, etc. than another set of ambiguous characters of another key and its corresponding mapped character.

As mentioned above, FIG. 5 illustrates an embodiment where four sets of ambiguous letters (e.g., of keys 15201-15204) and their corresponding mapping letters (shown on the bottom of each key) are on the eight keys 15201-15208 of the keypad 15200.

The same principle can be applied to entering other (ambiguous) characters of a key. For example, to enter an arabic numeral (e.g., 0-9), the user may press the corresponding (first) key (15207 in this embodiment) and (preferably simultaneously) press the corresponding second key.

If the number of keys of a keypad is insufficient to map each ambiguous character of a key to a different key of the keypad, more than one character may be mapped to a key such that at least each of some of the mapped characters for a key may be assigned to a different interaction with the key.

It will be apparent that the process of inputting precise letters as described above may be used with embodiments of the input and/or correction processes of the predictive data entry system of the present invention described in this and the earlier parent patent applications filed by the present inventor to reduce and/or eliminate the need for speech.

It must be noted that the keypad of this embodiment is shown as an example only to illustrate the principles just described for entering precise letters. Other embodiments may be contemplated by those skilled in the art.

As described above, when a word is input using the word prediction system of the present invention (e.g., when initial input information (e.g., key presses, and speech if provided) corresponding to a word is finished being input), if the word recommended by the system is not a desired word, the user can perform a correction process by providing additional input information corresponding to the desired word.

According to one embodiment of the correction process of the present invention:

a) if the user provides a predefined (correction) command without providing speech, such as providing a predefined pressing action for the correction key, the system preferably provides the word with the next highest priority (e.g. based on frequency of use) among the groups of words in the dictionary database of the system corresponding to the initial input information provided by the user. Each additional command (e.g. pressing the modifier key) without providing speech may preferably and separately provide another word of the group having the next highest priority.

b) If the user provides a predefined command, such as providing a predefined pressing action for a correction key, and provides speech corresponding to at least one of the characters of the word, preferably the starting character, the system preferably provides the word with the highest priority (e.g. based on frequency of use) among the words in the dictionary database of the system corresponding to the combined (e.g. initial and additional (e.g. correction)) input information provided by the user. In case no speech is provided, each additional command (e.g. pressing a modifier key) may preferably and separately provide another word having the next highest priority among the plurality of words in the database corresponding to the initial and the additional input information provided by the user.

As the description of this embodiment continues, if it is desired that a word and a word recommended by the system based on initial input information have the same starting character, then the user may provide additional input information by providing (e.g., pressing against a modifier key if predefined requirements, and providing) speech corresponding to (e.g., speaking) the character, in order to navigate within a plurality of words in the dictionary corresponding to the initial information and having the (e.g., starting) character.

The system may recommend a word (e.g., preferably, the word having the next highest priority) from among a plurality of words in a dictionary database of the system that correspond to the combined (e.g., initial and additional) input information provided by the user. In the case where no speech is provided, each additional command (e.g. pressing a correction key) may preferably provide another word having the next highest priority among the words in the word database corresponding to the initial and the additional input information provided by the user.

It has to be noted that according to one approach, the additional input information (e.g. speech) provided during the modification process may replace the corresponding prior (e.g. initial and/or) speech information provided by the user.

Different methods of entering precise characters without providing speech have been described previously. According to another embodiment of the invention, the invention may be designed such that when a user presses (e.g. and preferably holds) a first ambiguous key corresponding to/representing a desired (ambiguous) character (e.g. a letter), the system may provide a number of (soft) keys (areas on the screen of the device) such that each of the plurality of (ambiguous) characters of the ambiguous key is assigned to each of said (soft) keys. The user may then press the corresponding (soft) key to enter the precise (unambiguous) character.

FIG. 5a shows, as an embodiment, an electronic device 15310 having a data entry system of the present invention that uses several keys 15301-15308 that each replicate the keys 15201-15208 of the keypad 15200 of FIG. 5, in which a user can press (e.g., and hold) the key 15304 to unambiguously enter the letter "d". The system may show at least six on-screen keys 15309, each representing/corresponding to one of the multiple letters represented by/assigned to the key 15304 (e.g., there may be more on-screen keys to represent other characters represented by the key 15304). The user may then press the on-screen key corresponding to the letter "d".

It will be apparent that the exact letters entered by this embodiment may be used in the process of entering words by the predictive data entry system of the invention.

According to one method of the above embodiment, each time the user presses an ambiguous key (e.g., in this example, a hard key), the system provides a corresponding non-ambiguous key (e.g., a soft key/on screen key). In this case:

-if the user provides a pressing action for a corresponding non-ambiguous (soft) key

-the system provides the corresponding unambiguous character

If not (e.g. if the user provides a pressing action for a key other than the on-screen key related to the ambiguous key pressed by the user)

-the system takes into account said ambiguous key press action provided by the user

The user may provide one or more successive pressing actions (e.g., providing successive nonsense characters) on a plurality of nonsense (e.g., mapping) keys corresponding to one ambiguous key pressing action.

If the ambiguous keys are separated on opposite sides of the device, preferably a pressing action on each of the ambiguous keys may preferably provide a corresponding non-ambiguous key (on screen key) on the opposite side of the device as shown in fig. 5 a.

The device utilizing the data input system of the present invention may be a communication device using a dial-up scheme. In this case, as mentioned in the prior patent application filed by the present inventor, the keypad may have at least ten keys which may comprise the data input keys of the present invention, wherein (e.g. when the device is not in a data/text input mode) each of the ten keys may represent one of the arabic numerals 0 to 9, such that a pressing action on each of said keys (without speaking) provides the corresponding arabic numeral to permit dialing of a (telephone) number. As shown in the embodiment of fig. 5a, in addition to the keys 15301-15308, the keypad of the device 15310 includes two additional keys 153011, 153012 for duplicating the numeric keys of a telephone-type keypad so that (e.g., when the device is not in text mode) the ten keys may permit dialing of numbers.

It should be appreciated that to switch between the data/text input mode and the dialing mode, a switching device (e.g., a button, voice command, etc.) may be used.

As noted in the various patent applications filed by the present inventors, the system may use one or more of any speech recognition systems to recognize the user's speech. For example, the recognition system may be based on the user's voice, lip/face movement during speech, lip relationships during speech, bone/skin vibrations during speech, and the like. It is obvious that for such identification corresponding instruments (e.g. microphones, cameras, sensors, etc.) can be used.

Different methods of presenting the predicted words during the correction process of the present invention have been described in different patent applications filed by the present inventors. According to another approach, during/after entering the correction process, the starting character of the predicted word and/or the (each) precision input character/letter may be presented in a predetermined manner, such as by one or more predetermined characters (e.g., "_", "-", etc.). After confirming the word, the system may enter the word (e.g., with a complete character) by a predetermined confirmation means, such as, for example, assuming a space character is provided.

For example, as shown in fig. 6, by considering a keypad 16000 and a portion 16009 of the system dictionary, in order to input the word "crown", the system may recommend the word "crown" after the user presses the corresponding key. The user can press the correction key without providing voice. The system may provide a character set ". rowd", where the character "indicates the starting (e.g., predetermined) exact character (e.g.," c "in this case) of the word. Thus, by considering the same fig. 6, in order to enter the word "naomi", the system might recommend the word "crown" after the user presses the corresponding key. The user may press the correction key and speak the letter "n". The system may recommend the word "naked" (e.g., or a character set "saked", where the character "indicates the exact character of the word (e.g.," n "in this case)). Next, the user can press the correction key without providing the voice. The system may recommend a character set "× omic", where the character "×" represents the starting (e.g., predetermined) exact character of the word (e.g., "n" in this case).

As previously mentioned, other precise characters of the recommended word may be shown in the same manner.

As previously described, according to one embodiment of the present invention, during the correction process, the user may provide (e.g., by speaking) more utterances than one exact character of the word. According to one embodiment of the present invention, when the system recommends a word (e.g., a character string including characters representing exact characters), if the user speaks a letter, the voice may correspond to the next ambiguous letter of the recommended word (e.g., the character string) in a predefined manner.

According to another embodiment of the present invention, during the correction process of the present invention, if the words corresponding to the input information provided by the user have the same starting character (e.g., except for the exact character), the system may replace the (e.g., starting) character (e.g., character string) of the recommended word with the predetermined character (e.g., ") of the recommended word. For example, in fig. 6, by considering a keypad 16000 and a portion 16009 of a plurality of words of a system dictionary, in order to input a word "crown", the system recommends the word "crown" after the user presses a corresponding key. The user can press the correction key without providing voice. The system would understand that the intended word starts with the letter "c" and would recommend a character set ". x.d", where the character "x.x" represents the starting character starting with the letter "c" and having more than three similar characters, multiple words (e.g., in this case, becoming the exact characters of the two words "crown" and "crown" starting with the character set "crown").

According to one embodiment of the invention, the user may provide a predetermined interaction, such as a predefined pressing action for the (same or different) correction key and speak a character, wherein the speech may correspond to a (next) first non-similar character corresponding to the input information of the user and having several words of the same starting character. In the above embodiments, where such interaction with such a modifier key is provided and a letter (e.g., "d") is spoken, the system will understand that the speech corresponds to the last letter of one of the words "crown" or "crown" (e.g., in this case, preferably, the word "crown" because the word "crown" has a higher priority and may preferably have been shown to the user earlier). According to one method of presenting a predicted word, such characters may flash on the screen to let the user know the location of the character (e.g., within the desired word) of the desired word to be spoken.

According to one embodiment, pressing a key 16002, 16001, 16002 may correspond to the following words starting with the same letter "t", by considering a portion 16008 of the database:

-that

-text

-tear

the system may first recommend the word "that". To enter the word "tear", the user may press the correction key without speaking. The system may show the word "text" while at the same time flashing the letter "a" to indicate to the user: if one wishes to enter the next letter starting with the letter "t", he preferably speaks the third letter of the desired word because the preceding characters of the desired word are all similar to the currently presented word. In this embodiment, the user may press a predefined modifier key and speak the letter "x" to cause the system to present/select the word "text".

According to another embodiment of the invention (of the correction method), the starting character of the predicted word and/or the (each) precision input character/letter may be presented separately from the ambiguous character of the word. In the above embodiments, as the keys 16002, 16001, 16002 are pressed, it may correspond to the following words starting with the same letter "t":

-that

-text

-tear

the system would first recommend the word "that". To enter the word "text", the user may press the correction key without speaking. The system understands that it is desirable that a word should start with the letter "t" (it is predefined that t becomes the exact letter of the word). The system may select the next word (e.g., the word "text") corresponding to the user's input information and show it as two separate parts "t" and "ext". The separation method may be a predefined separation method that is provided in advance, such as providing a separator (e.g., "_") or the portions may be in different rows, have different colors, etc. The user may confirm that the combined portion is the desired word by entering a confirmation process (such as a space symbol) (e.g., or other action such as providing an end-of-word signal).

As previously described, different methods of inputting precise characters, such as a multi-tap system, may be provided for the combined (e.g., including word prediction) data input system of the present invention. The multi-tap method may replace and/or be part of the system in addition to the methods described in the inventor's different patent applications, such as user speech for inputting precise characters.

According to one embodiment of the present invention, one or more "mode" keys may be used to enter precise characters (e.g., such as a single character, or such as a character during the entry of a word). By providing a predefined interaction with a (predefined) key, such as pressing and holding the "mode" key, and providing multiple keystrokes corresponding to a desired character (e.g., letter) on the corresponding (e.g., ambiguous) key, the desired letter can be accurately entered. For example, by considering fig. 7, to enter the letter "f", the user may press and hold one/any of the plurality of "mode 2" keys of keypad 17000 and press key 17002 (three (consecutive) times (e.g., a multi-tap system) (e.g., predefined to correspond to the location of the letter "f" (e.g., in this case, left to right) among other characters of the key) (e.g., a multi-tap system for entering the exact character as known to those skilled in the art).

It will be apparent that the precise characters entered by the multi-tap system described above may be well suited for use in all of the input and correction methods of the word predictive data input system of the present invention, to replace the use of other methods such as voice, full (mini) keyboard, or handwriting methods, or to be used in combination with other methods described above to enter precise characters in the combined data input system of the present invention. For example, by considering the keypad and system dictionary (part of the system dictionary) shown in fig. 7, to enter the word "fear", after entering the letter "f" through the multi-tap process as described above, the user may release the "mode 2" key 17002 (e.g., remove his finger from the key) and (single) press the key corresponding to the remaining (e.g., ambiguous) character (e.g., "ear") of the word "fear". By considering the exact starting letter "f", and the key presses corresponding to the remaining characters, the system provides the word "fear" having the highest priority among the plurality of words corresponding to the input information of the user. It should be appreciated that in order to distinguish between the mode (e.g., "amend") keys (e.g., "17007, 17008) corresponding to the precise characters of the input ambiguous keys (e.g., 17001-. This also permits the system to distinguish between the correction method of the present invention and the multi-tap process of the present invention. As an example, if the user wishes to enter the word "real", the system may recommend the word "real" after providing input information (e.g., a key press action to enter the word "real" as described). The user may then press the mode/"modifier" key (e.g., 17007) once (without speaking) and the system will recommend the next word corresponding to the user's information, which in the embodiment of fig. 7 is the word "feat".

It is apparent that the predefined (e.g., key) interactions in combination with the multi-tap actions that inform the system of the use of the multi-tap action in order to input a precise character may be of any kind, such as another type of press/swipe action for the corresponding mode key, voice or handwriting commands, through interaction with a full (miniaturized) keyboard, and so forth.

In the multi-tap method for inputting precise letters as previously described, if two consecutive characters to be input are not on the same key, the user may press and hold the first key (e.g., the "mode 2" key) and provide a corresponding multi-tap action on the key corresponding to the first desired character and continue to hold the first key (e.g., the same "mode 2" key) in the pressed and held position (e.g., without releasing the first key) followed by providing a multi-tap action on the key corresponding to the second desired character.

On the other hand, if two consecutive characters to be entered are on the same key, the user may provide separate presses and hold actions for the first ("mode 2") key and corresponding multiple presses of the key for the corresponding character for each of the characters.

As previously mentioned, the combined (e.g., including word prediction) data entry system of the present invention may be provided with different methods of entering precise characters. The method may be an alternative to and/or be part of the system, in addition to the method for inputting precise characters, such as user speech, described in the inventor's different patent applications.

For example, when a user presses and holds a key (e.g., for at least a predefined time delay), typically after a predefined time delay between every two characters, the characters/letters on the key are recommended to the user one by one (e.g., shown on a screen) after the predefined time delay between the characters. As an example, by considering fig. 8, to enter the word "rear," the user may press and hold key 17102. After a predetermined time delay, the system will understand that the user wishes to provide a long press action so that it begins recommending/showing the on-key characters (e.g., letters) to the user one by one based on a predefined order (e.g., according to this/one embodiment (e.g., left to right), "a" then "b" then "f," … …, etc.). When the system reaches the letter "r" and recommends it (e.g., shown on the screen), the user may release the key 17102. The system will select the letter "r". Next, the user may tap/press a key corresponding to the remaining letter "ear" of the word "real" (e.g., then finally provide a word end signal such as entering a space character). By considering the input information provided by the user, the system selects the word "real" which, in this embodiment, is the only word in dictionary 17109 that corresponds to the input information.

The same method can be used to enter exactly any character on/of any key. For example, by considering fig. 8, to enter a character ". times.on a key 17107, the user may press and hold the corresponding key 17107. The system recommends the characters "+", then "_", then "#". At this point, the user may release the key 17107. The system enters the letter. In the embodiment of fig. 8, the same rule may be applied to any one of the input arabic numerals 0 to 9, and the like.

The same method can be used for character entry when the "mode" key is utilized. In the embodiment of fig. 8, for example, to enter the character "F" (e.g., capitalization), the user may press a mode key (e.g., 17107) and press and hold the key 17102. The system recommends the letters on the keys as described above, but in their upper case. When the system recommends the letter "F", the user may release the key 17102 (and/or 17107). The system may enter the letter "F".

It must be noted that in any method for inputting precise characters, such as the multi-tap and/or long press described, the recommended order of the characters may be dynamically defined by the user during data entry. For example, the user may inform the system that he wants the system to recommend the corresponding character on a key from left to right, or vice versa, when he provides a long press action on the desired key. For this purpose, the user may provide predefined interactions, such as additional interactions (e.g., keys, voice, etc.) accompanying the long key press action. For example, in the above-described embodiments, a single press and hold action (e.g., for a key corresponding to a desired character) may inform the system that the corresponding characters on the key are recommended one by one from left to right, while a double press and hold action (e.g., on a key corresponding to a desired character) may inform the system that the corresponding characters on the key are recommended one by one from right to left. This is advantageous when a desired letter on a key (e.g., the letter "w" on the key 17101) is near the end of the list of letters on the key, in the above-described embodiment, if the user provides a double press and hold action on the key 17101, the user may reverse the predefined recommendation order (e.g., "z", then "x", then "q", … …, etc.) to enter the letter "w" faster. According to another embodiment, the presentation order may be based on the priority (e.g. from highest to lowest) of the letters of the keys/keys in relation to each other.

The method of inputting precise characters/letters is described and preferably used with the word prediction data input system of the present invention as described in this and other patent applications filed by the present inventors is by way of example only. It should be understood that other methods may be used for the same purpose. For example, according to one approach, instead of/in addition to a press-and-hold action for a corresponding key for inputting a desired character, the user may use a slide-and-hold action (e.g., for the corresponding key). According to one embodiment, the left-to-right swipe and hold action for the corresponding key may cause the system to recommend (e.g., show on the screen) the corresponding characters (characters of the key) one by one starting from the left character to the right character, while the right-to-left swipe and hold action for the corresponding key may cause the system to recommend the characters one by one starting from the right character to the left character.

It has to be noted that although different types of interactions have been described for entering the exact character in each set of characters of a key with several sets of characters (e.g. lower case, upper case), it has to be noted that said various types of interactions have been shown and described as embodiments for entering said set of exact characters. It is obvious that the general idea is to use different types of interaction for each set of characters on a key (for entering precise characters) so that the system knows which set of characters the user's interaction is intended for.

Additionally, appending a different type of press and hold action (e.g., a press and hold action with a key) may inform the system of a character that repeats a key. For example, the user may provide a predefined press (e.g., and eventually hold) action for a first key (e.g., a mode key) and press and hold a key (e.g., a second key) corresponding to a desired character to cause the system to recommend the corresponding characters on the key one by one (e.g., on a screen) (as described previously). At this point, when the system recommends (e.g., prints on the screen) the desired character, the user may, for example, release the first (e.g., mode) key while continuing to hold the second key in the pressed position to inform the system to repeat the letter (e.g., on the screen). The system may repeat the character until the user releases the second key. At this point, the system may stop repeating the character.

It must be noted that it is apparent that any word (e.g., inclusive or non-inclusive) within a system dictionary may be entered by entering its characters precisely character-by-character in accordance with a precise character entry method such as that described above.

As mentioned in the different patent applications filed by the present inventors, the data entry system of the present invention may be integrated in a device with a retractable flexible (e.g. plastic) display, wherein according to one embodiment of the present invention, the plurality of keys in the keypad of the present invention may be separated into two groups and integrated in a device with the display unit, and wherein each key in the group of keys is on a different side (e.g. on opposite sides) of the extended display/device when the device/display is in the retracted position. It is apparent that the display may be retracted in any manner/form, such as folded, slid, etc. Thus, it may be expanded in any manner/form (e.g., expanded, slid open, etc.). For successive embodiments utilizing different methods for entering precise letters (e.g., during entry of a word) by the combined (e.g., including predictive) data entry system of the present invention, it must be noted that the precise letters can be entered at any location within the word. For example, based on the keypad 17200 and the dictionary 17209 of FIG. 9, to enter the word "text," a user may first press the keys 17202 and 17201 corresponding to the initial letter "te" of the word "text. He may then provide a long press (e.g., press and hold)) action on key 17202. The system may show (e.g., flow show) a group of characters on the keys one by one (e.g., show letters on the keys one by one) to the user. For example, the system may show letters (e.g., a, b, m, t, r, x, f) one by one (beginning) based on their priorities (see table 17210) relative to each other. The user may release the key when the system shows the letter "x" (e.g., in this embodiment, the letter "f" may not be shown because the user stopped the flow at "x"). This informs the system that the letter "x" must be entered exactly at that location (e.g., third) within the desired word. Next, the user can press/tap on the key 17202. By taking into account the input information provided by the user, the system may enter the word "text", which in this embodiment is a unique word corresponding to the user's input information.

As previously mentioned, long press motions for entering precise letters may also be used in the correction process of the present invention instead of using speech, or as an additional means of entering information (e.g., the system may also include presses and utterances for entering precise letters) to provide a correction process when speech is not desired. For example, after the user provides (e.g., initial) input information (e.g., key presses, and final speech (if desired/needed in the case that the system also has)), for inputting a word, if the system recommends an undesired word, the user presses and holds a predefined key, such as a correction key (e.g., as described in various patent applications filed by the present inventor, the press and hold action with a key may be a predefined interaction, such as a single/double press/tap/swipe action for the key and holding the key in a pressed position (e.g., without removing the finger/stylus from the key for at least a predetermined time delay), or a press (e.g., and hold) action for a first predefined key (as described) and a predefined press and hold action for the (e.g., second) key (e.g., as described). These events have been described in different patent applications filed by the present inventors). The system may recommend one word (e.g., one word for each different starting character) one by one among a plurality of words in the system dictionary corresponding to the input information of the user and starting with different characters in the word group of the key corresponding to the first character of the desired word, wherein preferably, the word having the highest priority among the plurality of words in the system dictionary corresponding to the input information of the user and starting with the (different) character.

When the system shows a word starting with the same letter/character as the desired word, the user releases the (correction) key. The system recommends the word. If the word is not the desired word, the user may provide a predefined (e.g., single) press/tap action for a predefined key (e.g., preferably a correction key). The system may show/recommend a next word (e.g., a word starting with the same character as the recommended word) corresponding to the initial, and additional, input information of the user, and wherein the next word preferably has the next highest priority among the plurality of words in the system dictionary corresponding to the combined (initial and additional) user input information (e.g., the process of recommending words based on the output and additional input information provided by the user has been described in detail in the description of the modification process of the press and talk data input system of the present invention).

On the one hand, if the new recommended word is still not the desired word, the user may repeat the press/tap action (e.g. not hold) for the correction key (e.g. or another predefined key). The system recommends a word corresponding to the combined information, wherein the word has the next highest priority with respect to the previously recommended word or the like.

On the other hand, if the recommended word is the desired word, the user may notify the system by a predetermined interaction such as providing a word end signal (e.g., entering a space character).

It must be noted that, during the correction process, when the user provides a press-and-hold (e.g., long press) action for a predetermined (e.g., "correct") key, instead of recommending words one by one as described, the system may recommend a plurality of characters of the key corresponding to the starting character of the desired word one by one. After the user selects the exact character, the system may recommend the/the corresponding word (e.g., based on the principles described above).

With long press actions for entering precise characters during the text input and/or correction process, it is perfectly integrated within the (predictive) data input system of the present invention that utilizes pressing and speaking. This combination provides a unique system. The long press action for inputting precise letters without speaking does not conflict with the flick action required for precise character input requiring speech. Thus, their combination may not change the behavior of the user when using the user's voice and the system together. The user can enter characters or words with or without speech at any time during the text entry without being forced to switch between the two separate systems/modes. As an example, on the one hand, the user may enter the exact letters of a word included in the system dictionary using a long press method (e.g., without speaking), and on the other hand, he may enter another word not included in the system dictionary, e.g., character by character, using the press (e.g., tap) and spoken data entry system of the present invention.

It has to be noted that even a speaker-free iterative process can be integrated within this combined "speech/no-speech" system without changing the basic interactions with the inventive (e.g. predictive) data input system combining press-and-talk and long press actions for inputting precise characters or words, such as:

the user may press (e.g. and hold) a first predefined key, such as the/a "mode" key, and provide a press-and-hold (e.g. long press) action on a second key corresponding to the desired character (e.g. the character the user wishes to repeat), as described before, in order to select the exact letter, and then providing a predefined interaction, such as releasing the first key, may inform the system that the user wishes to repeat the exact letter. The system repeats the letter until the user releases the second key. If the system is in "capitalization locked" mode, the/the characters may be repeated in capitalization (e.g., "MMMMM"), if not, they may be repeated in lowercase (e.g., "MMMMM").

It should be noted that, alternatively, the user may continue to use the iterative process used in conjunction with the press and talk method for the data input system as described in detail previously.

The system may include a feature for permitting a user to decide the speed at which characters are presented to the user (e.g., on a screen) during a long press (e.g., press and hold) process. For example, a novice user may wish to display characters slower (one by one), while a skilled user may want to display characters faster.

As an example of a preferred implementation, by considering FIG. 9, for example, a user may provide at least the following tasks (or additional tasks not listed below) (e.g., any/some):

- according to a methodTo enter the word "feat" without speaking, the user may press and hold key 17202 until the letter "f" is recommended to the user. At this point, the user may release the key 17202 and then tap on the keys 17201, 17202 corresponding to the remaining letters of the word "feat". The system may recommend the word "fear" corresponding to the user's information (e.g., initial information). The user may then press/tap (once) on the "correct" key. The system may recommend the word corresponding to the user's initial information and having the next highest priority. The word is the word "feat". The system recommends the word and the user confirms it.

- According to a further method of the above-mentioned type,to enter the word "feat," the user may first press a key (e.g., initial information) corresponding to the letter of the word. The system may recommend the word "that" having the highest priority among the plurality of words corresponding to the input information of the user. The user may then press and hold the correction key. The system may recommend to the user a plurality of words in the dictionary that correspond to the user's initial information one by one, and wherein each of the words starts with a different letter of the first key pressed during the initial input information, and wherein each word has a highest priority among the plurality of words that have the same input information and start with the same character (e.g., this event is described above). In this embodiment, the words may be "bear", "gear", "real", and "that". When the system shows the word "gear" (e.g., starting with "f"), the user may release the correction key. The word is selected but this is not the desired word. At this point, the user may provide a single press/tap action on the correction key. The system may show the next word corresponding to the user's initial information and having the same starting letter as the word "fear", and wherein the next word has the next highest priority after the word "fear". The characterIs the word "feat". The system recommends the word and the user confirms it, for example by entering a space character.

- In order to enter a word outside the dictionary,such as the word "zoro," the user may input the word character by pressing a key corresponding to each letter of the word and speaking the corresponding letter.

- In order to repeat the letter "m",the user may press and hold a predefined key, such as the correction/mode key, and press and hold key 17202 until the letter "m" is recommended to the user.

At this point, it is predefined that the user may release the correction/mode key while continuing to hold key 17202 in the pressed position. The system repeats the letter "m" until the user releases the key 17202.

- In order to repeat the letter "m",the user may press and hold key 17202 and speak the letter "m". The system repeats the letter "m" until the user releases the key 17202.

It must be noted that although it is preferred to combine non-speech methods that integrate the input of precise letters with speech-on-demand methods (for the same purpose) so that the user does not have to switch between modes, it is obvious that the system can be designed such that those methods are not implemented on the same data input system, and therefore the user can use the switching means to switch from one system of the invention to another press and talk data input system (e.g. from (e.g. predictive)), to predictive (e.g. with long press actions)/multi-tap data input system, and vice versa.

It has to be noted that the order in which characters are recommended to the user based on multi-tap/long press as described can be predefined differently. For example:

it may be based on the degree of priority of the corresponding characters in relation to each other (see, e.g., table 17210 of fig. 9), or:

it may be based on the priority of the corresponding word at the beginning of the long press action, or:

-alphabetical order, or:

-and the like.

Presenting the letters to the user may employ a class-looping process such that after the last corresponding character is recommended to the user, if the user has not yet selected a letter (e.g., and continues to tap multiple times or long press the key), the system begins showing the character again starting with the starting character.

According to one approach, the system may include a feature for permitting a user to reverse the order of presentation of recommended letters. For example, if the user wishes to reverse the sequence, he may press (e.g., and ultimately hold) another key, such as a backspace key, before/with providing the multi-tap/long-press action.

It must be noted that, with respect to the press-and-hold action, the multi-tap action may also be used with the correction process of the system. So as not to conflict with the "modify" key function, the multi-tap function for selecting a desired letter (e.g., or a word starting with the desired letter) for a press-and-hold action as described may use another predefined key (e.g., a space/backspace key, or another alphabetic key) in combination with a key representing the desired precise character, with a predefined definition. For example, to enter the letter "m," the user may press (e.g., finally hold) key 17205 and tap a key a number of times on key 17202 until the recommended word "m.

As previously described in the various patent applications filed by the present inventors, the word prediction data entry system of the present invention may also use other data entry methods, such as a full keyboard (e.g., at least the alphabetic keys are unambiguous), in addition to or instead of user speech, to accurately enter at least one of the characters of the word to be entered (the remaining characters of the word may be entered by pressing the (ambiguous) keys of the keypad of the data entry system of the present invention, as described above). The (full) keyboard may be an on-screen keyboard, or any kind of physical keyboard, such as a QWERTY keyboard.

It has to be noted that although the above described method/device may be an inconvenient system requiring many keys, there are many (types of) devices on the market which already comprise or are to comprise such a full keyboard. One example of such a device includes a mobile device having a miniaturized and/or on-screen full keyboard, where the use of such a keyboard for long and dense text entry is impractical.

The data entry system of the present invention may be integrated within such a device to permit the keyboard to be used in conjunction with the keypad of the present invention to more easily and quickly enter data (text) (e.g., if desired/designed not to use voice). For example, during entry of a word by the word prediction data entry system of the present invention, a user may accurately enter one or more (beginning, middle, or end) characters of the word using the full keyboard and press an ambiguous key (an ambiguous key of the keypad of the present invention) corresponding to the remaining characters of the word (e.g., initial input information). Also, for example, if the system recommended word (e.g., the word recommended based on the initial information) is not the word that the user wishes to enter, the user may enter one or more (e.g., starting) characters of the desired word (e.g., additional input information) during the correction process, for example, by, for example, providing a predefined interaction such as pressing (e.g., and holding) the correction key to inform the system of the correction process, and pressing the corresponding key of the full keyboard.

As mentioned above, such an input method may replicate the use of the (e.g., corresponding ambiguous keys and) phonetic and correction procedures for entering precise characters during the entry of words by the predictive data entry system of the present invention. The remaining input and correction procedures/principles (e.g., using correction/next alone to recommend next word) may still be as described previously. For example, after the one or more characters are entered during the correction process, if the system still recommends an undesired word, the user may provide a (e.g., separate) press action on the/a correction key to cause the system to recommend the next word corresponding to the user's combined (e.g., initial and additional) input information. It will be apparent that the full keyboard may be used in addition to or instead of methods such as speech for entering precise characters

One such device 17230 is shown in FIG. 10 as an example, where the two sets of keys 17217 and 17218 of the data input system of the present invention (in this example, the keys 17231 & 17238 are separated into two sets of keys 17201 & 17208, such as those replicating the keypad 17200 of FIG. 9) may include the alphabetic keys of the present invention (four ambiguous keys) in this example to be combined with the nonsense keyboard 17219 of the QWERTY keyboard of the device 17230 for entering text/data. For example, by considering the dictionary 17239, to enter the word "ping", the user may press the non-ambiguous key 17215 of the keyboard 17219 (providing the letter "p") and then press the (ambiguous) keys 17234, 17234, 17233 of the device/keypad of the present invention corresponding to the remaining character "ink" of the word "ping". By considering the first letter "p" of the precise input, and based on the (ambiguous) key presses 17234, 17234, 17233, the system may recommend the word "pick" in the system dictionary having the highest priority among the words corresponding to the input information. The user may provide a predefined (e.g., key) interaction, such as pressing a predefined modifier key (e.g., 17238). The system may recommend the next word in the dictionary that corresponds to the user's input information, where the word corresponds to the input information and preferably has the next highest priority relative to the last recommended word. The word is the word "ping". The user can confirm the word.

According to another embodiment, to enter the word "sink", the user may press the (ambiguous) keys 17233, 17234, 17234, 17233 of the device/keypad of the present invention corresponding to the characters of the word. By taking into account the (ambiguous) key press, the system may recommend the word "king" with the highest priority among the plurality of words in the system dictionary corresponding to the input information. The user may provide predefined (e.g., key) interactions (e.g., notify the system of the revision process), such as pressing a predefined revision key (e.g., 17238), and pressing keys 17215, 17216 of the keypad 17219 that correspond to the two exact starting letters of the word "pu" in the desired word "punk". The system may recommend the next word in the dictionary corresponding to the user's input information, wherein the word preferably has the highest priority relative to other words in the dictionary corresponding to the combined input information. In this embodiment, there is only one word corresponding to the user's input information. The word is the word "punk". The user can confirm the word.

It must also be noted that the full keyboard may also be divided into multiple (e.g., two) portions and positioned on the sides of the device such that each portion is adjacent to a corresponding thumb of the user. The keyboard is an on-screen keyboard and thus each of the portions may be on a respective side (adjacent a respective thumb) of the screen of the device. Preferably, the keys of a full keyboard may be small, while the keys of the keypad of the present invention may be large. Also, it is preferable that the keys (separated/not separated) of the keyboard may be arranged such that the key representing the most frequently used character/letter is close to the thumb of the user.

The system may also use any other means for entering precise letters, which may be any type of keyboard (e.g., chordal keyboard (half-QWERTY, Fitaly, shrinkkey board (e.g., fastap keypad), telephone-type keypad, etc.) and have a predefined interaction process set for the keyboard/keypad for entering precise single characters.

It must be noted that the arrangement of characters/letters on the multiple (e.g., four) keys of the keypad of the present invention may be arranged in a manner to be based on the multiple words of the system dictionary, so that most words of text provided by the system may be provided without the need for entering precise characters or correction processes.

Although a full keyboard may include punctuation characters and PC commands/functions, as described in the various patent applications filed by the present inventors, the system's keypad may include at least punctuation characters and PC commands/functions (e.g., at least those that do not require speech so that when the system/user is not using speech they can also input them through the keypad without speaking).

The principle of such a combined full keyboard and ambiguous keypad system may be based on the use of the keyboard as little as possible.

It must be noted that although in many paragraphs the keypad of the present invention is shown in a split form factor, it is apparent that the keys of the keypad may be, for example, in close proximity to one another to permit use of the keypad with the fingers of a single hand.

The system may use some keys of (the keyboard of) the device to execute/copy at least some of the keys of the keypad of the invention. For example, in FIG. 10, the keys 17231 and 17238 may be detachable/engageable keys or they may be part of a (full) keyboard of the device. Also, the device may have additional keys (e.g., 17227, 17228) that may be used for the same purpose (e.g., duplicating at least some of the keys of the keypad of the present invention) in addition to the keys of the full keyboard.

It has to be noted that it is obvious that at any time during the text input by the data input system of the invention combining a (full) keyboard and a keypad of the invention, the user may write a whole word (e.g. exactly character by character) with the full keyboard if he so wishes.

According to one embodiment of the invention, if the (full) keyboard used for entering the precise characters is a virtual (e.g. on-screen) keyboard (with smaller keys), during the entry of a word by the word prediction system of the invention, when the user presses a key of the virtual keyboard to enter the precise letter of the word, it may happen that his finger touches more than one key. In this case, the system may consider each of the one or more keys that were pressed/touched and provide input information for each of the keys and the remaining key presses of the user-provided word (e.g., input information including the precise key). For example, if a user's finger touches two keys while touching the full keyboard and provides a ambiguous key press corresponding to the remaining characters of the word, the system considers two input information, each comprising one of the characters. Thus, according to one embodiment, the system may recommend a word having the highest priority among a plurality of words corresponding to either of the two input information.

Different approaches to multi-tap systems have been described previously. As previously described, and with reference to fig. 8, according to one approach, a user may press (e.g., and hold) a first key (mode 2 key, e.g., 17103) and provide a multiple press action on the key corresponding to the desired character. This event has been described previously.

According to another approach, the user may first press (and hold) the user may press (e.g., and hold) a first key (the key corresponding to the desired character) and provide a predefined corresponding multi-press action on a second key (e.g., the mode 2 key). The first press (and hold) action may be a predefined press action, such as a single/double press/slip (and hold) action or the like.

According to another method, to enter an exact character, the user may press (and hold) a key corresponding to the desired character. The system may map a predefined number of characters (e.g., letters) on the key to a number of other keys (e.g., other keys of a keypad). At this point, the user may provide a pressing action on the key to which the desired character has been mapped. If there is only one character mapped to the key, it is selected, but if there is more than one character mapped to the key, the system maps the characters to different keys (which may include the same key). The user may then press the key to which the desired character is mapped. If the pressed key contains only one character that maps the key, it is selected, but if there is more than one character that maps to the key, the process is repeated until the desired character is the only character that maps to a key, so that the user selects it by pressing the key.

It should be appreciated that according to one approach, after the first mapping action, if there is more than one character mapped to the key, the user may select the desired word by providing a multiple press action on the mapping key corresponding to the desired letter. The predetermined number of press actions may be based on different parameters, such as the priority of letters associated with other letters mapped to the key, etc. This event has already been described in detail.

It has to be noted that in many paragraphs two keys have been pressed (e.g. pressing (and holding) a first key and pressing another key) for describing the predefined interaction, it has to be noted that it is obvious that a second key is pressed next (or simultaneously) after first pressing any of said keys.

As for devices including full keyboards, there are also devices (other than telephones) including keypads with other keys. The additional keys may be used to provide/duplicate at least a portion of the keypad of the present invention, and the telephone-type keypad may be used for multi-tap or mapping purposes for inputting precise characters.

Different methods of entering precise characters without providing speech (e.g. during input of words by the predictive data entry system of the invention) have been described previously. As previously described, according to one embodiment of the present invention, the keypad of the present invention may be designed such that a user may press (e.g., and preferably hold) a first predetermined key (e.g., preferably, any other key (such as an alphabetic key) than the mode key may be considered as the first key, which is referred to as a mode 2 key in fig. 11 to be distinguished from the mode key) and press a second (e.g., ambiguous) key corresponding to a desired character (e.g., letter), and the system may map at least some of the characters (e.g., such as letters) of the (ambiguous) key onto different keys of the keypad in a predetermined sense. The mapping process may be such that each key to which the character is mapped contains a small number of the characters (e.g., preferably one character). The user may then provide a predefined press action on the map key corresponding to the desired character to enter the character. Also, for example, if the key to which the desired character is mapped contains more than one mapped character, the user may, for example, provide a multi-tap/shuffle (e.g., single, double, etc.) process to select the desired character in order to enter the desired character.

According to one embodiment of the invention, when the user provides a press (e.g., and hold) action on a first key and provides a press action on a second key (e.g., an ambiguous key corresponding to the desired character), the system may not map one of the characters (e.g., such as the one with the highest priority among the other characters on the key). If the user releases the first key without pressing any map key, the character may be selected by default.

Fig. 11 illustrates a keypad 17300 of the present invention as an embodiment in which characters of english/roman are assigned to four keys 17301 to 17304 thereof. For example, if the user wishes to enter the letter "e," he may, for example, press (and hold) key 17304 (e.g., the mode 2 key when pressed and held) and (single) press key 17301 corresponding to the letter "e. Next, as shown in the embodiment of FIG. 11a, the system may map the letters assigned to the keys/key interactions (e.g., in this embodiment, except for the letter "e" with the highest priority among the letters) to some of the keys 17301, 17303, 17305, 17307, 11309 of the keypad. At this point, if the user releases key 17304, the system provides the letter "e" by default. According to another embodiment, to enter another letter, such as the letter "w" (e.g., while pressing and still holding the key 17304 in the pressed position), the user may press the corresponding key 17303 to which the letter is mapped.

As previously mentioned, the system may map more than one character to a key. For example, the key 17302 may include the seven letters "ABFMRTX", and a specific character "-". As an embodiment, if the user presses (and holds) a "mode 2" key (such as key 17301) and provides a (single) pressing action on key 17302, the system may map the letters and specific characters (e.g., in this embodiment, except for the letter "a" with the highest priority among the letters) to some of the keys 17302, 17304, 17306, 17308, 11310 of the keypad, as shown in the embodiment of fig. 11 b. To enter the letter "t", as an example, the user then presses key 17310 a single time. Also, as an embodiment, in order to input the letter "x" or a specific character "-", the user may double-press the corresponding key 17310 or 17302, respectively.

As previously mentioned, in many paragraphs it is apparent that the principles described above for inputting precise characters may be used with the word prediction data input and correction embodiments of the present invention. As an example, to enter the word "test," the user may enter the letter "t" precisely as described above. He can then release the key 17301 and provide a single pressing action of each (ambiguous) key corresponding to the remaining character "est" of the word "test". Based on the input information provided by the user, the system recommends the word "test" and the user can confirm it (for example, providing a single press action on each (ambiguous) key corresponding to the character of the word "test" would provide the word "best" with a higher priority).

The method of inputting precise characters may also be used in the correction process of the present invention. For example, if the user provides a single press action on each (ambiguous) key corresponding to the character of the word "test", the system would recommend the word "test" with the highest priority associated with the key press. At this point, the user may for example provide a pressing (e.g. and holding) action on a corresponding key of a predefined type (e.g. on the correction key), and if desired, he may provide a (single) pressing action on the key corresponding to the first character of the desired word and follow the mapping process as described above to select the desired letter "t" (e.g. if the user only informs the system of the first exact letter of the word, he may not have to provide again a pressing action on the key corresponding to the starting letter of the desired word during the correction process, upon which the system may enter the mapping process). Based on the initial information (e.g., ambiguous key presses) and additional information (e.g., precise letters) provided during the correction process, the system may provide the word "test".

If the user wishes to refine/enter additional precise characters of a word during the correction process, such as the second, third, etc. character of the desired word's starting character, then according to one method, for example, after selecting/refining the first character without releasing the press and holding the key, he may (single) press on the key corresponding to the next (e.g., second) character of the word and follow the mapping process to select/refine the second character and do the same for the third character, etc.

It is apparent that when more than one character is mapped to a key, the character with the highest frequency of use among the mapped characters can be assigned to the simplest interaction, such as a single press action on the key (e.g., and a double press for the next highest priority character, etc.).

The principles explained above can be applied to the precise input of other (ambiguous) characters of a key. For example, by considering fig. 11, to enter one of the arabic numerals (e.g., 0-9), the user may press and hold the corresponding (first) key (e.g., 17303 in this embodiment) and press the corresponding second key 17308 (preferably simultaneously). Next, as shown in FIG. 11c, the system may map the Arabic numerals 0-9 on different keys (in this embodiment, on five keys). Next, (while keeping the key 17303 in press and hold actions), the user may, for example, provide a single press on the desired key to enter a first digit on the key and/or provide a double press to enter a second digit on the key.

It should be noted that in many cases, such as when the keypad is virtually and/or theoretically (e.g., dynamically) and/or physically separated into two portions (e.g., positioned on opposite sides of the device as described in the patent applications filed by the present inventor), the user may utilize portions of the keypad with different thumbs, and thus, the characters of the ambiguous keys of an interaction (e.g., by a first digit of the user) included in a first portion may be mapped to the keys of a second portion (e.g., by another digit of the user).

It has to be noted that instead of a press and hold action, the user may provide another predefined interaction, such as a gliding action. These events have been described in different patent applications filed by the present inventors.

It must be noted that the examples given above are provided to illustrate the principles of the implementation. It is apparent that the mapping process may include any predefined type of interaction with at least one target, such as a key, and wherein characters may be mapped to any number of keys. These events have been described in detail in different patent applications filed by the present inventors.

As previously mentioned, it must also be noted that a full keyboard (for entering precise characters) for use with the (predictive) data entry system of the present invention may be divided into at least (preferably) two parts and positioned on either side of the device so that each part is adjacent to the user's thumb. If the keyboard is an on-screen keyboard, each of the portions may be on a respective side (adjacent a respective thumb) of the device screen. Preferably, the keys of a full keyboard may be small, while the keys of the keypad of the present invention may be large enough to be easily manipulated by a user. Fig. 12 illustrates a full (e.g., QWERTY) keyboard as an embodiment, separated into two portions 17411, 17412 and positioned on multiple sides of a corresponding device (e.g., UMPC/PDA/smartphone/etc.) and combined/used with the (predictive) data entry system of the present invention and its keypad (e.g., replicating keypad 17200 of fig. 9) (e.g., also separated into two portions 17401, 17402 on both sides of device 17410). In this embodiment, the screen 17409 of the device is positioned between the separate keys. The user may hold the device and (simultaneously) manipulate (keys of) the device.

It must be noted that although the full keyboard is referred to as a QWERTY keyboard, it is apparent that it may be any type of keyboard, such as a full keyboard of the "AZERTY" keyboard (french) version), a keyboard in which the alphabetic keys used are generally closer to the user's fingers, and the like. Furthermore, the ambiguous keypad of the present invention can be of any type and kind, with any number of keys (e.g., a telephone-type keypad), configured with letters on the ambiguous keys, etc. These events have been described in detail in different patent applications filed by the present inventors.

It is apparent that the separate full keyboard is used during/together with the (principles of the) predictive data entry system of the invention to replace the need for user speech for entering precise characters or as an additional means of entering precise characters. The principles of the data input system (e.g., data input, correction, etc.) of the present invention for inputting a word based on providing ambiguous (e.g., by several keys) key presses and non-ambiguous characters corresponding to the word (e.g., by different means, such as at least one of speech, full keyboard, handwriting/recognition, multi-tap, mapping process, etc.) have been described in detail in different patent applications filed by the present inventors. It will be apparent to those skilled in the art that any other method for providing the ambiguous information and the non-ambiguous characters described may be used with the principles of the data entry system of the present invention as described above.

According to one embodiment of the invention, at least some (e.g., ambiguous) keys of the inventive data entry system (e.g., letter keys assigned to at least, for example, four keys) may use some keys of a full keyboard (e.g., QWERTY) for entering precise characters during entry of words by the inventive (predictive) data entry system. This means that the same key can be used both for entering the exact character corresponding to the desired word and for entering the ambiguous key press corresponding to the desired word during the input/correction method of the invention. This may permit, for example, the use of keys of existing devices, as well as permitting the provision of smaller new devices.

Fig. 13 illustrates a mobile device 17500 with a smaller QWERTY keyboard 17509 as an embodiment. The apparatus may include the predictive data entry system of the invention which may use the keyboard to enter precise characters during the entry of words. In addition, some of the keys 17501, 17508 of the keyboard 17509 may also be used with the predictive data entry system of the present invention to replicate the (e.g., ambiguous) keypad of the present invention as shown and described in many embodiments. At any time (e.g., as during the entry of a word), the predefined (e.g., key) interaction may inform the system to treat the key as a key of the keyboard or as a key of the keypad.

According to one embodiment of the present invention, the combination keyboard and keypad system is used with the data entry system of the present invention (e.g., by default) when the user uploads the system. Preferably, at the outset, the keys 17501 through 17508 are defaulted to operate as the keypads of the present invention (e.g., keys 17521 through 17528, respectively, of the keypad model 17502 of the present invention) unless otherwise instructed by the user.

As an example, to illustrate some of the principles described above, by considering FIG. 13, to enter the word "seat," the user may first press key 17510 of the full keyboard to indicate to the system that the desired word must start with the letter "s". The user may then press the keys 17501, 17502 of the (ambiguous) keypad of the present invention. By considering the corresponding portion 175018 of the database of the present invention and based on the input information (and priority) provided by the user, the system can recommend the word "seat" and the user can confirm it.

The user may also enter the same word, for example, by pressing the keys 17503, 17501, 17502 of the keypad of the present invention. The user may then provide a predefined interaction, such as providing a predefined key interaction, to switch the keys 17501-. For example, the user may press (e.g., and hold) a predefined key (e.g., such as key 17505 of a keypad, or key 175011 of a full keyboard) to inform the system that the toggle action is provided, and press key 17502 to cause the letter "t" corresponding to the last character of the desired word "seat" to be entered accurately. Based on the input information provided by the user, the system may recommend the word "seat".

It must be noted that the user may provide a predefined interaction such as a predefined key interaction (e.g., in the above-described embodiment, releasing a depressed and held key) to switch the key 17501 and 17508 back to a key that functions as the (ambiguous) keypad of the present invention.

According to another embodiment, to enter to "seat", the user may first press the (ambiguous) key 17503, 17501, 17502 corresponding to the character of the word. The system will provide the word "year" corresponding to the key press and having the highest priority. At this point, the user may provide a predefined interaction, such as providing a predefined press action on a predefined key, to inform the system to enter the corrective process (e.g., the press action may be a predefined interaction such as a single press action, a press and hold action, a double press action, etc.. this event has been described in detail). The user may then provide the exact letter corresponding to the (e.g., start) character via the full keyboard (e.g., by pressing the corresponding key of the full keyboard), and if desired, provide a correction key to navigate (e.g., based on priority) within a plurality of words corresponding to the initial and additional information, etc., provided by the user (the correction process of the present invention has been described in detail previously).

In this embodiment, after entering the correction process, the system may preferably automatically switch to full keyboard mode, after entering the correction mode, the user may press key 17510 to inform the system that the desired word must start with the letter "s". Based on the initial key press and additional information provided by the user, the system may recommend the word "seat".

It is apparent that in the above embodiments, if a word is not included in the dictionary of the system, the word may be entered character by character through the full keyboard.

The principles of the data entry system using a full keyboard described above may also be applied to any other type of keyboard/keypad (e.g. using its corresponding system for entering precise characters) and application of the (ambiguous) keypad and (e.g. predictive) data entry system of the present invention.

According to one embodiment of the invention, at least some (e.g., ambiguous) keys (e.g., letter keys assigned to at least, for example, four keys) of the (predictive) data entry system of the present invention may use some keys of any (e.g., existing) device, such as some keys of a telephone-type keypad of a device. As previously mentioned, the (predictive) data entry system of the present invention may also use the input system (e.g., multi-tap) of other (e.g., telephone-type) keypads used for entering precise characters.

Figure 14 illustrates a mobile device 17600 with a telephone-type keypad 176010 as an example. The data entry system of the present invention may use some of the keys (e.g., 17601-17608) of the telephone-type keypad to, for example, respectively replicate the keys 17501-17508 of the keypad 17500 of the present invention shown in FIG. 13. The data entry system of the present invention may also use a multi-tap method for use with a telephone-type keypad for entering precise characters, such that a desired word is entered by the predictive data entry system of the present invention during entry of the word or during a correction process at least one of a plurality of precise characters/letters of the word. The principle of common keys between the keys of the telephone type keypad and the keys of the keypad of the present invention can be used for full keyboards as well.

As an example for illustrating some of the above principles, by considering the keypad of FIG. 14 and the dictionary 175018 of FIG. 13, to enter the word "seat," a user may first press (e.g., and hold) a predefined key 17607 or 176011 (e.g., to enter a telephone-type keyboard mode) followed by four quick press actions (e.g., multi-tap) on the key 17603 corresponding to the letter "s" on that key to accurately enter the letter "s" (e.g., the key 17603 corresponding to a key having the letter "PQRS," where the user must quickly press four times to enter the letter "s". the principles of the multi-tap method used with a telephone-type keypad are known to those skilled in the art). The user can then switch the ambiguous keypad of the present invention, for example, by releasing the hold key. The user may then (mono) tap on the key 17601, 17602 corresponding to the (e.g. ambiguous) remaining character of the desired word "seat". Based on the user's precise (e.g., character) and ambiguous key information, the system provides the corresponding word with the highest priority. The word is the word "seat". The system may recommend it and the user may confirm it. It will be apparent that in the above embodiments, when the system is in the telephone-type keypad mode, the arrangement/distribution of the letters on the keys is considered to be similar to the standard arrangement of roman/english letters on a telephone-type keypad.

It is apparent that during input/correction of a desired word by the predictive data entry system of the present invention (e.g., also using the keypad of the present invention), a telephone-type keypad using a multi-tap method (e.g., or other (known) methods) may be utilized to replicate the foregoing, and that during predictive data entry/correction of the present invention also using the keypad of the present invention, all principles of the predictive data entry/correction system of the present invention for entry of precise characters may be utilized to enable entry of the precise characters.

It must be noted that if a word is not included in the dictionary of the system, the word may be entered character by a multi-tap system (e.g., telephone type keypad and the like). It is obvious that the words can be added to the system database automatically/manually.

It has to be noted that entering said mode informing the system that the exact character entered during a mode belongs to the correction process may be provided by any predefined interaction such as one of the following:

1. providing a push-hold action on a key (e.g., key down)

2. Providing another type of interaction such as predefined press/swipe actions (such as single/double press/swipe actions on predefined keys)

3. And the like.

The user may exit the above mode (e.g., after entering the exact letter) in the following manner:

1. releasing the hold key (e.g., the key is raised)

2. Again providing said another type of interaction such as a predefined press/swipe action (such as a single/double press/swipe action on said or another predefined key)

3. And so on.

As previously shown in fig. 12, a full (e.g., QWERTY) keyboard for use with the data entry system of the present invention may be divided into multiple (e.g., two) portions and positioned on different sides of the corresponding device. FIG. 14a shows expandable apparatus 17620 with first portion 17621 and second portion 17622 in a closed position. The device further comprises an expandable flexible display unit. Devices of the type described have been described in different patent applications filed by the present inventors. The keypad of the present invention having several keys may be divided into two sets of keys 17623 and 17624, and each set of the keys may be positioned on one of the portions of the device. Also, the full keyboard may be divided into two groups 17625 and 17626, and each group of the keys may be located on one of the portions of the device. Figure 14b shows the device and its display 17627 in an extended position. The display may be made of a flexible material, such as plastic, to be rolled up (e.g., within the device), wrapped (e.g., around the device), etc., when the device is in the closed position.

Also, as an example, FIG. 14c shows an expandable apparatus 17610 with an expandable display unit in a closed/retracted position. The device further comprises an expandable flexible display unit. Here, the keypad of the present invention having several keys may be divided into two groups of keys 17611 and 17612 (e.g., in the form of two multipurpose keys) and positioned on opposite sides of the device. The device also includes a full keyboard 17613 that may be used with the data entry system of the present invention. The device may also have another display unit 17614, a microphone 17616, a speaker 17615, etc., and include telephone functionality. Fig. 14d shows the device and its extendable display 17618 in an extended position (e.g., extending from the longer side of the device in this embodiment). The display may be made of a flexible material, such as plastic, to be rolled up (e.g., within the device), wrapped (e.g., around the device), etc., when the display is in a retracted position, which may be equipped/made in such a way as to remain in a free-standing flat position.

The keys of the keypad and the keys of the full keyboard of the present invention may both be disposed on the device so as to coincide with the extendable display unit 17618 when the device is in the closed position. The keys of the keypad of the present invention (e.g., in this embodiment, each multi-purpose key may replicate five keys (four in the corners and one in the center)) and/or some of the keys of the full keyboard may replicate the keys of a telephone-type keypad, such that they may operate as telephone-type keypads for dialing when the device is in the closed/telephone position. In such a case, the telephone-type keypad may preferably fit on (smaller) display 17614 (e.g., in the direction of the display) according to one approach.

The (word) predictive data entry system of the invention, which utilizes a full keyboard, may be packaged such that a separate data entry unit is formed to interact with any type of (electronic) device, either wirelessly or by wire.

For example, as shown in fig. 15, the device 17700 can be used as an external (mobile) keyboard for an electronic device such as a computer 17701, GPS, mobile phone, etc. According to one embodiment, the device may transmit the user-provided key press (e.g., if used) to the corresponding electronic device and may perform processing within the electronic device. According to another method, the processing may be performed within the data input unit and the results may be sent to the corresponding electronic device. The data input unit may also include its own display unit to permit the user to view what he writes so that he may be remote from his computer.

As previously described and illustrated, the data entry system of the present invention may include a method for entering precise characters without speaking. One of the described methods is a mapping system where, when a user provides a predefined pressing action on an ambiguous key, the corresponding character on the key can be mapped to other soft/hard keys such that for each of the mapped keys only a few of the characters (e.g. preferably one character) are mapped such that the exact character is entered by a simple pressing action (e.g. single/double tap/swipe) on the mapped key.

According to one embodiment of the invention, the corresponding characters of a predefined pressing action on an ambiguous key may be mapped to several (e.g. dynamic) on-screen keys. This event has been described previously and is shown earlier in fig. 5a, and further examples of preferred embodiments are described hereafter.

By considering FIG. 16a, as an example, an electronic device 17910 has a data entry system of the present invention that uses several keys 17901 and 179010 that in this example respectively replicate the keys 17301 and 173010 of the keypad 17300 of FIG. 11. As previously described, for example, to unambiguously enter the letter "d," the user may first press (e.g., and ultimately hold) the key 17904. The system may show at least six on screen keys 17911, each representing/corresponding to one of the letters represented by the keys 17904 (e.g., there may be more on screen keys to represent other characters represented by the pressing action on the key 17504). The user may then press the on-screen key corresponding to the letter "d".

It will be apparent that the precise characters entered by this embodiment may be used during the entry of words by the predictive data entry system of the invention.

According to one method of the above embodiment, each time the user presses an ambiguous key (e.g., in this embodiment, a hard key), the system provides a corresponding non-ambiguous key (e.g., a soft key/on screen key). In this case:

-if the user provides a pressing action on the corresponding non-ambiguous mapping (on-screen) key:

-the system provides the corresponding unambiguous character

-if not, (for example, if the user provides a pressing action on a key other than the on-screen key, associated with the ambiguous key pressed by the user)

-the system takes into account said ambiguous key press action provided by the user.

According to one embodiment of the invention, if the characters of the predefined interaction with the ambiguous key are not part of a word, one of the characters cannot be mapped on the screen, so that if the user provides a corresponding pressing action on a key, but does not press any mapping key, the system presses the default to enter the character. As an example, as shown in fig. 16b, if the user provides a predefined press action on key 17907, the system may map the corresponding characters (except the characters ",") on the screen. The user may enter any of the mapping characters by providing a pressing action on the corresponding mapping key. If not (e.g., after the user presses another hard key), the system may enter a default character ",".

As previously mentioned, several consecutive pressing actions on a map key may provide several precise characters. For example, FIG. 16c shows the Arabic numerals 0-9 mapped on several on-screen keys after pressing key 17908. For example, to enter the number "1278", the user may press the key 17908, and then he may successively press the four mapping on-screen keys corresponding to the arabic numerals "1", "2", "7", "8".

Symbols assigned to mode (key) interactions can also be entered by this method. For example, a symbol (e.g., such as a lower case letter) may be assigned to a single press action on a first key (e.g., and mapped on the screen when the key is pressed), and the same letter in upper case may be assigned to a press (e.g., and held) on a mode key and a press on the first key. In this case, the system may map the symbol (e.g., capital letters) on the on-screen key (see the embodiment of FIG. 16 d), and the user may press the desired on-screen key.

The mapping method described above can also be used in the correction process of the present invention. For example, to input a word through the predictive data entry system of the present invention, after the user provides initial input information, if the system recommends that the word is not a desired word, then according to one method, the user may press a predefined key, such as a correction key. The system may map the characters corresponding to the first character corresponding to the desired word to provide an initial key press over several (dynamic) on-screen keys (e.g., preferably one character per on-screen key). The user may then press the on-screen key to inform the system about the first exact letter of the desired word. According to one approach, thereafter (e.g., thereupon), the system may map the character corresponding to the second character corresponding to the desired word to provide an initial key press over several (dynamic) on-screen keys. The user may then press the on-screen key to inform the system about the first exact letter of the desired word, etc., if desired.

Based on the initial input information and the additional input information, the system may recommend the corresponding word (e.g., based on priority).

It is apparent that, as previously described, according to one approach, after pressing the predefined (e.g., correction) key, the user cannot press any corresponding map key on the screen. In this case (e.g., after a predefined time delay or by providing a predefined command/interaction), the system may recommend the next word (e.g., based on frequency) corresponding to the input information (e.g., initial input information, and additional input information, if any) provided by the user (e.g., etc., if necessary).

The user may provide one or more successive pressing actions on the unambiguous key corresponding to one ambiguous key pressing action (e.g., to provide successive ambiguous characters).

If the ambiguous keys are separated on opposite sides of the device, it is preferred that each pressing action on the ambiguous key provides a corresponding non-ambiguous (on screen) key on opposite sides of the device (e.g., near the user's thumb).

As previously mentioned, the word prediction data input system of the present invention is based on the use of a small number (e.g., four) of keys to input ambiguous characters (e.g., and if desired) combined with additional information provided by other means, such as user speech or systems that provide precise characters (e.g., qwerty keyboards, multi-tap systems, handwriting, etc.) during or at the conclusion of word input (e.g., the correction process of the present invention). It should be appreciated that, in general, to enter an exact character during initial input information, a first predefined type of interaction may be used to enter the character, and during the correction process of the present invention, a second predefined type of interaction may be used to inform the system about the exact character (e.g., ambiguous keys (e.g., and non-ambiguous characters, if provided) of the word entered/recommended by the initial input information. For example, according to one embodiment, the system may be designed such that when a user wishes to enter an exact character when entering initial information for a word through the predictive data entry system of the present invention, he may press the mode/shift key and press the desired key on the (qwerty) keyboard (e.g., according to yet another embodiment, alternatively, he may double press the desired key on the (qwerty) keyboard). In this embodiment, for example, during the correction process of the present invention, a simple pressing action, such as a single pressing action on a desired key on a (qwerty) keyboard (e.g., without using a conversion/mode key), may correspond to providing the exact character (relevant information) of the word.

As previously mentioned, it is apparent that any other method or keyboard may be used for entering the precise character/information during the input of words by the predictive data entry/correction system of the present invention. For example, the keyboard may be a telephone-type keypad and the method may be a multi-tap system. FIG. 17 illustrates, as an embodiment, a communication device 18000 having a keypad of the present invention with six keys 18001-. In this embodiment, the exact characters during the input/correction of the word may be entered through the telephone-type key pad 18009, for example, using a multi-tap system (multi-tap systems are known to those skilled in the art). For example, to enter the word "bird" by considering a portion 18008 of the system dictionary, the user may provide two consecutive pressing actions (double-click) on the keys 18012 of the telephone-type keypad 18009 to accurately enter the letter "b", and then press the ambiguous keys 18004, 18002, 18004 corresponding to the remaining characters of the word "bird". The system provides the word "bird" by considering the (initial) input information (e.g., precise characters and ambiguous key presses). According to another embodiment, the user may first press the ambiguous keys 18002, 18004, 18002, 18004 corresponding to the characters of the word "bird". The system may recommend the word "turn" having the highest priority among the plurality of words corresponding to the key press (e.g., initial information). then, the user may double press a key 18012 (e.g., initial information) on the keypad 18009 to inform the system that the first character of the desired word must start with the letter "b". The system may recommend the word "bird" having the highest priority among the words corresponding to the initial and additional input information. As mentioned before, in order to distinguish between the entry of the exact letters/characters during the initial input of information and the entry of the exact letters/characters during the correction process, different predefined (key) interactions may be used. For example, during initial entry of information, precise characters may be entered by interacting with (e.g., multi-tap) corresponding keys of the telephone-type keypad without combining mode keys, while during the revision process, a user may, for example, press (e.g., and hold) predefined (e.g., mode) keys 18019 and interact with (e.g., multi-tap) corresponding keys of the telephone-type keypad (or vice versa).

It is apparent that the/the keypad/keyboard providing precise characters may have any number of keys and any type of alphabetic distribution. For example, letters of a language may be assigned to twelve keys on a telephone-type keypad, and the letters assigned to a key on the keypad may be designed such that a single pressing action on the key may provide the most common letters, a double pressing action on the key may provide the next most common letters, and so on.

(e.g., in the embodiments described above), a system combining user speech may be used instead of/in addition to a multi-tap system for entering the precise characters via ambiguous keys (e.g., those of a telephone-type keypad). According to one approach, a letter (e.g., preferably the most common letter) on an ambiguous key can be entered without providing a corresponding voice, e.g., by providing a (single) press on the key. Any other letter on the key can be entered by the pressing action and providing a corresponding voice (e.g., speaking the letter).

For example, by considering the phone-type key region 18009, a single pressing action on the key 18012 may correspond to the letter "a" without providing speech. To enter any other letters on the key (e.g., "b" and "c"), the user may single press on the key and speak the letter. This is particularly effective for obtaining better accuracy in the case of entering characters/letters through a telephone-type key region using the user's voice, wherein a first letter in each of two letters with ambiguous similar (e.g., voice-based) voice in a plurality of letters on a key (e.g., "e, d" on key 18013, "k, j" on key 18015, "m, n" on key 18016, "t, v" on key 18018) may be assigned to a (single) pressing action on the key without providing the voice, and a second letter in each of two letters with ambiguous similar voice in a plurality of letters on a key may be assigned to a (single) pressing action on the key and provide its predefined voice (e.g., say it). It is obvious that the corresponding assignment may thus be different if the speech recognition is lip based or lip & speech based.

According to one embodiment of the invention, a (predictive) data entry system benefiting from a full (e.g. qwerty) keyboard may be integrated with a (telecommunications) device in order to provide a multi-functional use of the device in an optimal way.

Fig. 18 shows a device such as a mobile phone/data device 18100 with a key set 18101, 18102 of the invention, for example to replicate the keys of the keypad of fig. 9. The device may be designed so as to extend in different directions for different corresponding functions.

Fig. 18a shows said device 18100 expanded in lateral position with keys 18101, 18102 of the invention, some of which (also) replicate a telephone-type keypad 18104, and a qwerty keyboard 18103 (for example for use with the predictive data entry system of the invention). The device 18100 may be designed so that it can be expanded and used in a landscape position when the user wishes to enter text, and in a portrait position when the user wishes to use it in a phone/dial mode, as shown in fig. 18 b. It is apparent that the display unit 18107 of the apparatus can thereby change the printing direction.

As previously mentioned, the predictive data entry/modification system of the present invention utilizing several (e.g., at least four) keys distributively assigned substantially all letters of a language may use any means for entering precise characters. As described above, any keyboard and method may be used to enter one or more precise characters/information about precise characters of a word during entry/correction of the word by the predictive data entry system of the invention. For example, the keyboard may be a telephone-type keypad and the method may be a push and talk data entry system of the present invention (where, for example, to enter precise letters, a user may push the corresponding keys and speak the letters). In the above data entry method, at least some other symbols (e.g., available on a PC keyboard) may be entered by interacting with at least the keys of a telephone-type keypad.

In embodiments where any of the total letters of a language are entered exactly by pressing the corresponding key and speaking the letter, other symbols such as:

a first predefined type of symbol, such as an arabic numeral assigned to a key, can be entered by a first predefined type of interaction, such as a single pressing of said key without providing speech

-a second predefined type of symbol, such as a punctuation character assigned to a key, can be entered by interacting with the first predefined type of said key and providing a predefined speech corresponding to said symbol

-a third predefined type of symbol, such as a function assigned to a key, can be input by interaction with the second predefined type of key, such as double pressing the key without providing speech

-a fourth predefined type of symbol, such as a command assigned to a key, may be input by interaction with the second predefined type of key, such as double pressing the key, and providing a predefined speech corresponding to the symbol

-and so on

-and vice versa

FIG. 19 illustrates, as an embodiment, a communication device 18200 having a keypad of the present invention, which in this embodiment has eight keys 18001 and 18008 that replicate the keys 18221 and 18228 of the exemplary keypad 18220 of the present invention. In this embodiment, the exact characters during the entry/correction of a word can be entered, for example, through the telephone-type keypad 18209 via the push and talk data entry system of the present invention. For example, by considering a portion 18230 of the system dictionary, to enter the word "fear," a user may first press a key 18213 of the telephone-type keypad 18209 and speak the letter "f" such that the letter "f" is entered accurately, and then press the ambiguous keys 18201, 18202 corresponding to the remaining character "ear" of the word "fear. By taking into account (initial) input information (e.g. precise characters and ambiguous key presses), the system provides the word "fear". According to another embodiment, the user may first press the ambiguous key 18202, 18201, 18202 (initial information) corresponding to the character of the word "gear". The system may recommend the word "that" having the highest priority among the plurality of words corresponding to the key press (e.g., initial information). According to one correction method, the user may then press (e.g., and hold) the correction key and press the key 18213 of the telephone-type keypad and speak the letter "f" (e.g., additional information) to inform the system that the first character of the desired word must start with the letter "f". The system recommends the word "fear" having the highest priority among the words corresponding to the initial and additional input information.

As mentioned before, in order to distinguish between the input of the exact letters/characters during the initial input of information and the input of the exact letters/characters during the correction process, different predefined (key) interactions may be used. This event has been described previously.

As previously described, according to another embodiment, one of a plurality of letters assigned to a key can be accurately input by pressing the key and speaking the letter, and any of the other letters assigned to the key can be accurately input by pressing the key and speaking the letter. In this case, as an embodiment:

different predefined type symbols, such as arabic numerals assigned to a key, can be input by another predefined type of interaction, such as pressing (and holding) a first key, and pressing the key alone without providing speech (or vice versa).

As previously described, in this embodiment, according to one method, a first letter of two letters having an ambiguity similar to a voice and assigned to a key (e.g., a predefined interaction with a key) can be input by providing the predefined interaction with the key without providing the voice, while a second letter of the two letters can be input by providing the predefined interaction with the key and providing a predefined voice corresponding to the letter (e.g., speaking the letter). It has to be noted that said similar speech may be based on any of at least the speech or lip movements used for providing (any of) said speech, e.g. used for speaking (any of) said letters.

It should be appreciated that in this embodiment, the assignment of one of several letters on a key to interact with the key without providing speech, and the assignment of other letters on the key to interact with the key and provide a predefined corresponding speech may be based on other principles. For example, according to one approach, the most common letters on a key (a on key 18012, "e" on key 18213, "i" on key 18214, "s" on key 18217, "t" on key 18218, "w" on key 18219, etc.) may be assigned to a (single) pressing action, such as a single pressing of the corresponding key without providing speech.

Preferably, if two (or more) letters/characters on a key have ambiguous similar voices in relation to each other, the letter assigned to the key (e.g., a predefined interaction with the key) and not providing a voice may be a letter used more frequently among the two (or more) letters. As a result, as an example, the letter "K" on key "5" of a telephone-type keypad, where the voice in english pronunciation would be ambiguously similar to the letter "J" on the same key (the letter "K" is more common than the letter "J") may be assigned to pressing the key without speaking. In this embodiment, each of the letters "J" and "L" (e.g., although the letter "L" is more common than the letter "K") may be assigned to press the key and speak the letter.

With continued reference to this embodiment, as an example, by considering fig. 19, to enter the word "ear," the user may enter initial and/or additional information as described in the previous embodiment. On the other hand, also as an example, in this embodiment, in order to input the word "alba", the user may first press the key 18213 of the telephone-type keypad 18209 without speaking to input the letter "a" precisely, and then press the ambiguous keys 18201, 18202 corresponding to the remaining character "lba" of the word "alba". By taking into account (initial) input information (e.g., precise characters and ambiguous key presses), the system provides the word "alba".

It is apparent that the precise letter (e.g., "a" in this embodiment) may (also) be entered during the correction process of the present invention based on the principles of the data entry and correction process of the present invention as previously described.

It is obvious that the/the keypad/keyboard providing the precise characters may have any number of keys and any type of letter distribution. For example, letters of a language may be assigned to twelve keys on a telephone-type keypad, and the letters assigned to a key on the keypad may be designed such that a single pressing action on the key may provide the most common letters, a double pressing action on the key may provide the next most common letters, and so on.

It has to be noted that although in different embodiments different keypads such as telephone type keypads or keypads with at least four keys, wherein at least substantially all letters of a language are assigned to the four keys, have been used to illustrate the principles of different embodiments of the present invention, it is apparent that any kind of keypad with any number of keys, wherein at least some of the keys of the keypad are/may be ambiguous, any type of keys, such as hard or soft keys, any configuration of symbols assigned to any predefined type of interaction with the keys of the keyboard, etc. may be used with embodiments of the combined (e.g. including predictive) data input system of the present invention. For example, if the predictive data entry system of the present invention uses a telephone-type keypad in english, then according to one embodiment, the user may use his eight ambiguous alphabetic keys to enter a word through the predictive data entry system of the present invention during which (e.g., during entry/correction) he may enter one or more precise characters/letters of the word by pressing the corresponding key and speaking (e.g., or other interaction such as pressing a first predefined (e.g., "mode/correction") key and pressing on the corresponding key multiple times to avoid using speech) the characters/letters. The user may also use one of the multiple (e.g., other) keys of the keypad for the revision process of the present invention, as described in this and other patent applications filed by the present inventors. It will be apparent that the modification method of the data entry system of the present invention can also be applied in conjunction with the use of a telephone-type keypad. For example, as described in this and other patent applications filed by the present inventors, according to one approach, a key (e.g., another) such as key "1" of a telephone-type keypad may be used as a "mode/modify" key and used in conjunction with the modification process of the present invention. As an example, by considering a telephone-type keypad similar to the telephone-type keypad 18009 of fig. 17, after pressing a key (e.g., initial input information) corresponding to the word "ball", the system may recommend the corresponding word "call" having a higher priority. As an example, the user may provide additional input information through the correction process of the present invention by pressing the predefined correction key 18019 and speaking/speaking the word "b". Next, the system may recommend the word "ball" corresponding to the initial and additional input information provided by the user. These events have been described in detail previously.

Although in many paragraphs describing the combined data input system of the present invention "keys" have been used to describe embodiments, it should be understood that other input means used by predefined types of interaction to provide different predefined input signals may be considered to replicate/replace keys and key interactions in embodiments of the present invention. For example, as described in this patent application filed by the present inventors and incorporated herein by reference, user interactions, such as finger or finger cover interactions and corresponding interactions, and other means such as styluses, gloves, etc., that replicate keys and key interactions, may be used in these embodiments.

It should be noted that in many paragraphs, the term "press" has been used as a class of examples of interaction with a key. The interaction may be any type of predefined key interaction, or any type of interaction with a key, such as touch/swipe/point, etc.

It must be noted that although in many paragraphs a keypad (e.g. or a few keys) has been used to illustrate an embodiment of the combined (e.g. including predictive) data entry system of the present invention, it should be appreciated that the keypad may typically have some ambiguous keys (e.g. ambiguous keys as already described hereinbefore).

In order to not repeat the principle of the data input system of the present invention frequently, in many paragraphs of the present application it is mentioned that one or more symbols, such as a character/word/part/function of a word, may be assigned to a key (e.g. or other input target in addition to a key). It should be understood that the symbols are generally intended to be assigned to the simplest interaction in a corresponding embodiment based on the principles of the data input system of the present invention as explained in many paragraphs.

Although in different embodiments of the invention it may have been mentioned that a voice recognition system (e.g. with a microphone) is used for sensing and recognizing the speech of the user, a lip reading system (e.g. with a camera) may be used instead of/or in addition to the voice recognition system for sensing and recognizing the speech of the user (or vice versa). As previously mentioned, the system may use different methods/means of detecting the user's speech, such as with means for detecting the user's speech provided by means such as the user's voice or lip movements. For example, an ear/bone microphone that detects a user's voice based on bone vibrations of the user's (e.g., ear) when providing speech may be used with the present system in addition to, or instead of, a conventional microphone (e.g., the principles of ear/bone microphones are known to those skilled in the art). It has to be noted that each of the means of detecting the speech of the user may be used separately or at least some of them may be combined.

Also, as previously mentioned, some or all of the methods of the data entry system of the present invention may use a linguistic regular text entry recognition system, such as the number of syllables of possible matching words, the number of words of possible recognition sentences, the position of words within phrases, a whole word system, and the like. These events are known to those skilled in the art.

It must be noted that in some paragraphs, the term "part-by-part" has been used to simplify the term "at least a portion of a word (e.g., by at least a portion of a word)".

It is noted that although for reasons of simplicity in many paragraphs the data input system of the present invention is described in terms of phrases such as "data input system of the present invention", "press/shuffle and speak data input system of the present invention", etc., it is to be understood that such phrases relate to the principle of the data input system of the present invention, as described in detail in many paragraphs, that is to consider the principle of a press/shuffle action according to several embodiments where it is possible to combine/combine with the speech information of a user, where the speech information indicates the presence or absence of a voice of the user (e.g., where the present system is typically used with a combined data input system of the present invention, such as during input of a word, at least one precise character of the word is input by the referenced system, and wherein at least one key press provided for entering the word is ambiguous). These events have been described in detail. Moreover, in many paragraphs, the data entry system of the present invention combining the press/shuffle and speak data entry system of the present invention and a word predictive data entry system (such as the word predictive data entry system of the present invention) has been recited in terms of phrases such as "the combined data entry system of the present invention".

It should be noted that although in many embodiments of the data entry system of the present invention, a sensitive surface such as a touch sensitive disk or touch screen has been used as an example, it should be understood that any other technique of detecting and analyzing user interaction with any surface may be used to define and/or use the regions/keys of the soft (e.g., dynamic) keypad. For example, as described above, the technique may be an optical detection technique or an IR technique that sets a virtual keypad (e.g., with a few keys/areas, where, for example, 4 keys/areas of the keypad are assigned at least substantially all letters of a language) on a (normal) surface, and detects a user's finger touching a key/area of the keypad.

Also, it must be noted that according to one approach, when a user enters a word using the predictive system of the combined data input system of the present invention, the system may print the most likely moment of the word corresponding to the input information provided by the user (e.g., before providing an end of word signal). It should be appreciated that the displayed word may change each time the user presses an additional key corresponding to the word. According to an embodiment of the present invention, unless a user provides a word end signal such as a space character at the time of ending the input information, the word prediction system of the present invention does not recommend an ambiguous word corresponding to information of the user (e.g., ambiguous key presses, and ambiguous information such as (key presses and) speech, handwriting, and the like).

It will be apparent that the principles of this invention and other inventions, in whole or in part, of their embodiments and/or their features, may be used separately or in combination. For example, at least a part of at least one word of a text may be entered by combining different methods or embodiments of the data entry system of the present invention. For example, a tap/swipe action on a sensitive surface as described in detail in this application filed by the present inventors and incorporated herein by reference (e.g., such as described in detail in this application filed by the present inventors and incorporated herein by reference, e.g., a tap and/or swipe action provided by a user's finger, pen, glove, etc., replicating interactions with a few keys, such as interactions with at least four keys of a keypad where at least one language letter is assigned to a few predefined keys, such as four keys of a plurality of keys (as described in detail in this application filed by the present inventors and incorporated herein by reference), the glove of the data input system of the present invention also replicates keys and different additional interactions, such as user speech, full keyboard, handwriting, and recognition, etc., for inputting precise characters) can be used with the predictive data input system of the present invention.

It must be noted that although in many embodiments the input information provided for entering a word is stated as being based on key presses, it should be appreciated that the user may provide other additional ways (such as voice, full keyboard keys, etc.) corresponding to the corresponding key presses and/or at least some of the characters of the desired word in addition to key presses (e.g., for better accuracy) if desired/permitted.

It must be noted that although in many paragraphs describing the data entry system of the present invention, the input information provided by the user is recited as characters that may correspond to words, it is apparent that, depending on the architecture of the system, in some cases the input information may correspond to a string of characters corresponding to words (see, for example, the partial word data entry system of the present invention). It will be apparent that the character data entry system of the present invention in combination with the partial word data entry system of the present invention may be used with the (e.g., word) predictive data entry system of the present invention, as described in detail previously. For example, to enter the word "action," the user may first press the keys corresponding to the letters "a" and "c" and then provide a swipe action on the key that is predefined to present a portion "station" of the word, and say that portion (e.g., "shen"). Based on the input information provided by the user, the predictive data entry system of the present invention may, for example, look for a word in which its two starting characters correspond to two key presses and the remaining characters of the word are the character string "station". The system will consider the word "action" and, depending on its frequency of use, the system will recommend the word to the user. In the above embodiments, the user may also enter the character "a" or "c" precisely by speaking either of the characters (e.g., while pressing the corresponding key) to help the system/user provide the desired word, for example, more quickly/accurately.

The use of the combined character and partial word data input system of the present invention and the predictive data input system of the present invention to input and/or modify/replace the words recommended by the system has been described in detail. As previously mentioned, such a system may include/cooperate with at least one data input system (such as an on-screen full keyboard or handwriting input/recognition system) to input precise characters that may correspond to at least one letter or portion of a word, instead of or in addition to a speech recognition system. The predictive data entry system of the present invention may take into account the precise character and additional input information, such as ambiguous key presses corresponding to the character of the word, to provide the desired word.

Also, for example, the combined data entry system of the present invention may be used with/implemented using an extensible computer (e.g., using the split keypad of the present invention), such as seen in fig. 70a through 70h and their corresponding description in PCT/US04/12082, incorporated herein.

It must also be noted that depending on the architecture of the system, partial words of one language may be included in a dictionary of a word stock of that language, or they may constitute a separate database and be used with the combined (e.g., including predictive) data entry system of the present invention.

It should be noted that a phonetic message corresponding to a letter is typically spoken of the letter.

It must also be noted that a key press corresponding to at least one character of a word generally refers to providing a sequence of key presses corresponding (sequentially) to the character.

It has to be noted that providing a predefined interaction with a key also means providing a press and hold action on another key and providing a press action on that key (or vice versa), as detailed above.

It must also be noted that although in many of the paragraphs describing the method/features of the invention key press information has been described as being used to input words, it is apparent that the method/features may also take into account other input information such as corresponding speech that may be provided by a user in addition to or instead of the key press information.

Thus, while there have been shown and described and pointed out fundamental novel features of the invention as applied to alternative embodiments thereof, it will be understood that various omissions and substitutions and changes in the form and details of the disclosed invention may be made by those skilled in the art without departing from the spirit of the invention. It is the intention, therefore, to be limited only as indicated by the scope of the claims appended hereto. It should be understood that the drawings are not necessarily drawn to scale, but that they are merely conceptual in nature.

In the correction process of the present invention, when the user presses the correction key without providing the voice, although it is preferable that the system may be designed to recommend the next word corresponding to the (e.g., combined) input information of the user and having the next highest priority, it is obvious that the system may be designed differently to provide the next word corresponding to the input information of the user and incorporating other principles (as follows)

-context of sentences/text, or

-alphabetical order, or

-others

It has to be noted that although in many paragraphs a pressing action on a key is recited as illustrating/exemplifying interaction with said key, it is obvious that said pressing action, as recited in a large number of embodiments, may be of any type, such as:

-single/double/multiple press/slide actions on said key

-simultaneously or not, a single/double/more pressing/sliding action on another key and a single/double/more pressing/sliding action on said key (or vice versa)

-single/double/more pressing/sliding and holding actions on another key, and single/double/more pressing/sliding actions on said key (or vice versa)

-and so on.

It must be noted that this combination of interactions can be very numerous and it is not possible to describe them in their entirety in order not to make the patent application lengthy. For the sake of brevity, it must be noted that such combinations may be apparent to those skilled in the art.

Furthermore, it is obvious that the interaction for entering/selecting the exact character during the input and/or correction process for entering the word may make the description of its possibilities possibly too long for the present patent application.

As an example, with the predictive data entry system of the present invention, by considering the components of FIG. 13, after the system recommends an undesired word to the user based on the initial input information provided, during the correction process for entering the desired word:

according to one approach, the user may press (e.g., and preferably hold) the correction key and enter the (starting) character/letter of the desired word by pressing the desired key of the full keyboard or by speaking the desired letter. The system may recommend words corresponding to combining the initial and additional input information. If the newly recommended word is still not the desired word, the user may provide a predefined interaction, such as a single press action on the modifier key, to scroll to the next word corresponding to the initial and additional input information, and if the next recommended word is still an undesired word, repeat the above process, etc.

According to another approach, the user may provide a predefined interaction, such as a predefined press action with at least one predefined key, such as a press and hold action on a first (e.g. mode) key, and press a desired key on the full keyboard or speak a desired letter to enter/select said letter. The system may recommend words corresponding to the combined output and additional input information provided by the user. If the newly recommended word is still not the desired word, the user may provide a predefined interaction, such as a predefined pressing action with/on a predefined key, such as a single pressing action on a correction key, to scroll to the next word corresponding to the initial and additional input information provided by the user, if the next word is not yet the desired word (e.g., pressing and holding the first/mode key simultaneously, if desired), and repeat the above process, etc.

It must also be noted that the combination of such interactions can be very numerous and that it is not possible to describe them in their entirety in order not to make the present patent application lengthy. For the sake of brevity, it must be noted that such combinations may be apparent to those skilled in the art.

It has to be noted that although the term "mode" (e.g. according to other terms) has been used to describe the interaction with a key, it is obvious that in general said (specific) term has been used to simplify the description of the corresponding embodiment. It is obvious that because the system uses a few keys, each key is typically used for more than one symbol (such as a character or function). Thus, for example, if the term "mode key" is used somewhere in this patent application, it generally refers to a key that can provide the function by a predefined interaction with the key.

According to one embodiment of the invention, the inventive press and talk (e.g. including word prediction) data input system utilizing speech recognition, such as at least both voice recognition and lip reading systems, may use the inventive keypad in such a way that characters, such as at least letters, assigned to a few (e.g. four) keys of the keypad are assigned to the keys in such a way that ambiguous similar speech related to each other is present by a first recognition method (e.g. voice) while letters/characters not having ambiguous similar speech (e.g. or vice versa) are assigned to the same key and/or to the same predetermined interaction with the key by a second recognition method (e.g. lip reading), so that when the user presses the corresponding key (or provides corresponding key interaction) and speaks any of the characters/letters, the system may be implemented by a system in said (speech) recognition systemAt leastOne to recognize the ambiguous letter/character. It is obvious that, as described in the different patent applications filed by the present inventors,based on the principles of the inventor's press and talk data input system, it may be preferable to assign characters/letters with ambiguous similar voices (e.g., in this embodiment, with both voice recognition (e.g., voice recognition, lip reading/recognition) methods) associated with each other to different keys and/or to different predefined interactions with the same key.

It must be noted that the word stock and the arrangement of the fields corresponding to each word illustrated in the different figures and their corresponding description are shown as examples only. It is apparent that the word stock of the present invention may have different fields and be arranged differently to, for example, enhance the searching and sorting of the word stock.

It has to be noted that the system for inputting precise characters is described only as an example for illustrating the principles of the predictive data input system of the invention (e.g. inputting a word based on pressing an ambiguous key and, if required, inputting at least one precise character of the word (information corresponding thereto)). It should be understood that numerous other methods and derivations based on the principles of the present invention and the illustrated method can be devised by those skilled in the art for entering precise characters and using the entered precise characters in the word input/correction process by the predictive data entry system of the present invention.

Claims (5)

1. A data entry system, the data entry system comprising:

a plurality of first type input devices, such as keys, for providing input signals to which symbols, such as letters of at least an alphabet of at least one language, are distributively assigned and such that at least two of the letters are assigned to at least one of the plurality of input devices, and in which pressing a key provides first input information corresponding to any of the letter ambiguities and second input information for supporting any of the following processes;

using a second type of input device for inputting non-ambiguous letters;

providing additional information corresponding to a letter assigned to the pressed key to assist the data entry system in selecting the letter among a plurality of symbols assigned to the input signal;

wherein the data entry system supports at least one word stock such that, to enter a desired word in the at least one word stock, a user of the data entry system can employ an input method selected from the group consisting of:

a first method comprising combining information corresponding to the word, wherein the combining information comprises providing the first input signal, such as a plurality of key presses corresponding to at least some of a plurality of characters of the word, and providing the second input information to accurately input at least one character, such as a letter of the word, by the input process; and

a second method comprising providing said input signal such as a plurality of key presses corresponding to substantially all characters of said word without providing said additional information.

2. The data entry system of claim 1, wherein the data entry system compares information provided by the method with corresponding information corresponding to a plurality of words in the word stock and recommends a corresponding word, and wherein, if the word recommended by the data entry system is not the user's desired word, the user can enter at least any of the following correction processes to provide further information corresponding to the desired word to the data entry system:

a first correction process, the first correction process comprising the steps of:

if the recommended word begins with the same letter as the starting letter of the desired word, the user presses a predefined key, such as the "next" key, and the data entry system recommends the next word in the word stock that corresponds to the combined input information provided by the user and that begins with the same letter as the recommended word;

if the recommended word does not start with the same letter as the starting letter of the desired word, the user providing information corresponding to the first character of the desired word to cause the data input system to identify the character and recommend a word that corresponds to the combined input information provided by the user and starts with the letter;

and a second correction process, the second correction process comprising the steps of:

the user provides information corresponding to a first character of a desired word, causes the data entry system to identify the character, and recommends a word corresponding to the combined input information provided by the user, wherein the word begins with the letter.

3. A data entry system according to claim 1, wherein the second type of input device is a standard PC keyboard or keypad in which pressing a key provides precise characters such as letters.

4. A data entry system according to claim 1, wherein said means for entering non-ambiguous letters is a handwriting recognition system.

5. The data entry system of claim 1, wherein the additional information is the spoken letter.