JP2006061632A - Emotion data supplying apparatus, psychology analyzer, and method for psychological analysis of telephone user - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an emotion data supplying apparatus which allows numerical evaluating the emotion and character of an individual to estimate its health, emotion, intelligence, psychological state, desire, and human relation, and taking the suitable response for the individual; a psychology analyzer, and a method for psychological analysis of a telephone user. <P>SOLUTION: The emotion data supplying apparatus comprises an operator terminal 3 for inputting an operator's answer information to a user's question information, and a psychology analyzer 1 which receives the user's question information from a user terminal 2 through a first circuit 4a, receives the operator's answer information from an operator terminal 3 through a second circuit 4b, and quantifies the psychological states of the user and the operator to analyze. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an emotion data providing apparatus, a psychological analysis apparatus, and a telephone user psychological analysis method that output analysis data obtained by quantifying a user's emotion and psychological state.

  For example, in the conventional image evaluation system, when evaluating the speech image of a language expression such as a name, the sound phase group such as the minimum pronunciation unit is set to “brightness (positive / negative)”, “strong / weak” ) ”, But it requires a significant survey. It takes emotion words from the language and classifies them into“ bright ”and“ strong ”and sounds such as the smallest pronunciation unit. We investigated where many of them were used and expressed their tendency numerically based on statistical methods.

  This image evaluation system is based on the classification and evaluation of emotional words in Japanese for many years, and has a ± B value (Brightness), which has a tendency of “brightness and darkness” with a sound base such as a minimum pronunciation unit, “ The ± H value (Hardness), which is a “strength” tendency, was obtained and organized (see Patent Document 1).

  However, the conventional image evaluation system is based on emotion analysis based on the classification and evaluation of emotion words, naming analysis of new products, name diagnosis of children and pets, personality diagnosis by name, game character parameter setting, and mail text The words were extracted to determine the sender's mood, and image evaluation was performed based on the current personality and emotions. However, this evaluation can be used to infer preference for each user by numerically determining individual emotions and personalities with higher accuracy. The system has been longing for.

There is also a need for an evaluation system that determines a user's emotion and personality through conversation (or communication) with the user and sends a message in accordance with the content of the user's remarks.
JP 2002-204153 A (see paragraph 0041, Table 1)

  The present invention numerically evaluates emotions and personalities that differ from person to person, infers the person's health, emotions, intelligence, mental state, desire, human relations, etc. , A psychological analysis device, and a telephone user psychological analysis method.

  According to an aspect of the present invention, for example, an operator terminal that inputs operator response information for user inquiry information, a user inquiry information from a user terminal through a first line, and an operator terminal from the operator terminal through a second line. And a psychological analysis device that receives the response information and digitizes and analyzes the psychological states of the user and the operator.

  One aspect of the present invention is, for example, a means for digitizing and analyzing a psychological state of a telephone user who makes a call through a telephone line, and a result of the analysis to a terminal of a group user whose telephone number is registered in an address book. The gist of the present invention is a psychological analysis device including a psychological analysis transmitting means for transmitting.

  In one aspect of the present invention, for example, the user's psychological state based on information on the five senses including vision, hearing, smell, touch, and taste is quantified and analyzed, and the user is based on the user history information stored in the user information management module. Psychologically, emotional numerical values including emotions are detected, matching solution information is output, and when numerical values that are depressed are detected, an encouragement message is output by letters, pictures, or sounds, and the user psychologically If it is uplifted, the gist is that it is a mobile phone equipped with a psychological analysis device that outputs a text or picture or the dream sense message by the five senses information or sound.

  One aspect of the present invention is, for example, a mobile phone that outputs emotion analysis data obtained by digitizing and analyzing first to third resonance frequency bands included in voice data of a telephone user to a display.

  In one embodiment of the present invention, for example, a telephone server automatically responds to a call from a telephone user's mobile phone via a mobile phone switch, stores voice data in a storage device, and reads voice data from the storage device for communication. Phone user psychology analysis that transfers to the application server through the line, analyzes and compares the voice data based on the emotion analysis data accumulated in the emotion analysis engine by the application server, and sends the matching emotion analysis data to the mobile phone through the mobile phone switch The gist is the method.

  One aspect of the present invention includes, for example, a microphone voice input step for inputting a voice of a telephone user to a mobile phone, a storage step for storing the voice input in the microphone voice input step in a memory, and a noise intensity mixed by the microphone voice input. An effective determination step for determining whether or not the voice is valid as emotion determination data, and a mail transmission step for transmitting the voice determined to be valid in the determination step to the server as an email attachment through a communication line; The server extracts parameters including the pitch, speed, and timing of the voice, analyzes and compares the voice based on the emotion analysis data stored in advance in the emotional analysis engine, and sends the matching emotion analysis data to the mobile phone through the communication line. Sending, and the mobile phone displays the emotion analysis data on the display graphic or text. Display step, and parameters including pitch, speed and timing of voice in mobile phone, and based on emotion analysis data stored in advance in emotion analysis engine, voice and five senses (viewing angle, hearing, smell, and The present invention is summarized as a telephone user psychological analysis method including a step of incorporating means for analyzing and comparing five senses information including taste) and displaying matching emotion analysis data on a display.

  When configured in this way, the words and phrases used by users, articles of interest, favorite stores, daily behavior patterns, how to spend time, etc. are analyzed based on biological theory, and each user is divided into colors, The direction of sensibility can be predicted from the human constitution, and the human mind closely related to the body can be evaluated numerically.

  According to the present invention, emotional data that can be individually dealt with by numerically evaluating different emotions and personalities for each individual, inferring the person's health, emotions, intelligence, mental state, desires, human relationships, etc. It is possible to provide a providing device, a psychological analysis device, and a psychological analysis method for any one of the occupation or character of the telephone user.

  The emotion analysis apparatus and psychological analysis apparatus in the following embodiments can use a human body color numerical evaluation system (hereinafter simply referred to as “evaluation system”). For example, the evaluation system includes an input unit that converts voice information, text information, aura check information, and the like into electrical signals.

  The human body color used in the present embodiment is generated, for example, at a stage where a static color as a base color closely connected to the human body and a transition to the next emotion in response to some event (event). It is a mind color based on the biological theory in the brain science including the dynamic color.

  If the input unit is a means for converting various information indicating the energy state of the human body into electrical signals that can be processed by the evaluation system, the keyboard, mouse, voice input device, image input device, communication line, human body, etc. Means such as a current measuring device, a human body static electricity measuring device, and an infrared measuring device may be used.

  The color dictionary receives an electrical signal from the input unit, digitizes it corresponding to a plurality of colors based on either the biological theory or the sound theory, or both the biological theory and the sound theory, and sends it to the hue calculation unit. Means for transmitting the color information can be used.

  The hue calculation unit adds a plurality of pieces of color information 12 to each other, and calculates and outputs a hue calculation result based on a numerical value corresponding to the color information. For example, a yellow binary number “101002” and an orange binary number “100301” are added and divided by a decimal number “2” as a numerical value corresponding to the color information of two colors, and a hue calculation result is output. Means can be used.

  The output unit connected to the hue calculation unit outputs hue system information corresponding to the static color and dynamic color of the human body based on the hue calculation result. This hue information is intended for multiple color information that can be inferred by human body parameters such as human health, emotions, psychology, intelligence, mental state, desire, and other human relationships, etc. by logical operation processing means. To do.

  For example, the evaluation system can numerically judge the person's emotion, psychology, and personality from the expression of words and human energy, and inferring preferences for each person, the words, services, environments, We can respond by providing products.

  Here, the biological theory is one of the constituent requirements of means for predicting the direction of completion from the constitution of the human body. Human psychology and emotions are closely linked to the body, so for example, an angry person has a mind color that the hormone that stimulates anger emotions for some event is the quickest secreted constitution. It is a systematic theory.

  In addition, it has been proved by biological theory that the constitution of an angry person has a high activity of liver cells in natural science and is liable to alcohol, but easily suffers from liver damage. Similarly, it is possible to construct a biological theory color dictionary in which at least five organs that induce joy, the liver is angry, the kidneys are afraid, the lungs are sad, the spleen is depressed, and the heart induces pleasure.

  In other words, fearful people have high kidney activity, sad people have high lung activity, sad people have high spleen activity, rejoicing people have high heart activity, hormones that stimulate these emotions It is also proved by biological theory that is the constitution that is secreted most quickly.

  For example, the balance of the body containing hormones can be estimated by using a measuring device such as the temperature of the acupuncture points on the palms and soles, the static electricity to be charged, and the current, and the body balance can be adjusted by stimulating the points. Proven in theory.

  The evaluation system can analyze the static color and dynamic color of the human body based on the biological theory by determining the internal energy state that appears from the emotions and actions of the human body. For example, it is possible to analyze a user's static color and dynamic color by inputting a word or phrase frequently used by an e-mail by a mobile phone user.

  In addition, based on the search tendency of web pages on the Internet, it is possible to identify the types of goods and services that the user is interested in, the type of store, etc., and analyze the user's static color and dynamic color.

  In addition, the user's daily behavior pattern and how to spend the day's time can be identified, and the user's static color and dynamic color can be analyzed. Therefore, the user's unique personality can be speculated and defined by two mind colors.

  Since humans have individual trends such as whether anger, sadness, melancholy, or joy comes first when the brain recognizes an event and moves to the next emotion, the evaluation system 1 uses the internal energy. The state can be analyzed by biological theory and expressed in different colors to estimate human emotions and constitutions.

  The mind-colored white person is neutral in all balances, so it is easy to maintain normality without being angry, angry or worried, or intimidated. The food and information that you want is also the nature of wanting something with as little stimulation as possible and a solid material.

  People with a mind-colored blue are often calm, benevolent and educated. This is because the hormone that stimulates sorrow and maternity is biased toward an easy constitution.

  Mind-colored red people have a pioneering spirit and a fighting spirit. This is because it relies on hormones that stimulate emotions of anger and propulsion.

  The mind-colored people mentioned above want food and information that suits their constitution, and when their constitution goes too far, they may act to accept food and information belonging to the opposite color even if it is painful. Many.

  This kind of behavior has become the main theory of brainologists in the field of brain science, as "the heart is in the brain". In this way, the personality of a person can be classified into, for example, 12 types of colors, and a system such as artificial intelligence can be made to learn the correspondence that matches the personality of each person.

  The numerical sensibility evaluation system (hereinafter simply referred to as “Kansei evaluation system”) is a fusion of a semantic network and a specialized program that individually correspond to the platform of customer products, and a sensibility inference unit, database, and artificial intelligence dialogue engine. Is intended for systems comprising

  The sensibility evaluation system inputs a text sentence from the dialog input unit, and performs a semantic analysis on the words of the text input via the artificial intelligence dialog engine. It is intended for a system that performs emotion analysis on the words that have been semantically analyzed by the semantic analysis unit.

  The Kansei evaluation system also includes a color dictionary based on biological theory, a tone dictionary based on tone theory, and a Kansei reasoning processing unit, and based on the color dictionary or tone dictionary, the sensitivity or emotion of words and psychology Intended for systems that analyze conditions.

  The sensibility reasoning calculation processing unit outputs an analysis result of a dynamic psychological state as color information expressing emotions such as anger, sadness, melancholy, and joy described above. For example, a means for outputting happy color information and sad color information to a database is targeted.

  For example, the database can be configured with a hard disk drive of about 1.4 megabytes that stores about 300,000 words and fixed phrases, but other storage devices may be used as long as they can store 300,000 words. .

  The database outputs the reply sentence extracted based on the analysis result of the dynamic psychological state to the dialogue output unit via the artificial intelligence dialogue engine. The reply sentence can be handled by text or voice, but this embodiment is intended for an interactive system using text.

  The text-based artificial intelligence dialogue engine is a learning-type dialogue-type artificial intelligence system that collects personal information while reflecting and reflects it in the conversation. For example, it is possible to execute text mining (morpheme analysis) processing on an e-mail, analyze the timbre of individual words with a lexicon, analyze emotions from sentences, and respond by dialogue.

  Here, the phonetic dictionary decomposes the speech expression that constitutes a predetermined language expression into the minimum pronunciation units expressed by one or both of consonants and vowels that are one of its basic components, It is possible to use a dictionary used for image evaluation in which one or a plurality of images having a minimum pronunciation unit obtained by analyzing a linguistic expression by a statistical method is quantified based on a tone theory.

  It also collects the user's own daily actions, answers to questions from the artificial intelligence dialogue engine, the tendency of the web being accessed, etc., classifies the user's personality with a color dictionary that uses biological theory, It is also possible to provide a system for changing the correspondence (for example, a mind color analysis system).

  In addition, we can systematize sound analysis and mindkara analysis to build a basic semantic network, and provide products that apply the human body color numerical evaluation system individually to the customer's platform that provides various services and products can do.

  For example, the human body color numerical evaluation system is installed in mobile phones, personal digital assistants (PDAs), digital home appliances, etc., and by providing the user's feelings using these products, accurate or individualized conversations can be provided. User emotions can be satisfied.

  A specialized semantic network that matches the target user group of a product that implements a human body color numerical evaluation system can be constructed on a computer server, and a conversation program can be customized to suit the product platform. In short, it is possible to easily provide specialized programs that match the planning of each platform.

  When the artificial intelligence dialogue using the biological theory of the present embodiment is illustrated, when “battle and it is easy to protect the victory” is input, a word is automatically cut out from the sentence. For example, words can be extracted from sentences using a word dictionary, and the color of each word can be determined.

  The input sentence corresponds to the color enclosed in parentheses for each word. For example, `` Battle (red) and (red) is easy (white) and wins (red and white) protects (green) is difficult (indigo) ''. The color can be determined.

  Similarly, if you enter a sentence that says “Do not forget important matters, do n’t forget important matters (blue)”, “Don't forget important matters (green), do n’t forget important matters (indigo)” (white), the color will be determined for each word.

  Also, in artificial intelligence dialogues using biological theory, the output part of artificial intelligence displays “Now, you ’re wearing clothes,” and recalls a color or color that matches the user's internal energy state. You can recommend clothes with the pattern you want.

  In this case, the biological theory determines the color of the “XXX” part that matches the basic constitution based on the static color of the human body and selects the corresponding word. Sentences excluding the “XXX” part are standard sentences and are stored in the output dictionary.

  In addition, if the user inputs the word “I want to do XXX this evening (because I feel like XXX)”, the biological theory is activated and the color of the time zone is analyzed from “Tonight”. The color of the action can be analyzed from “XX”, the color of desire can be determined from “I want to do”, the static color and the dynamic color can be calculated, and the current user color can be determined.

  Of course, as the fixed sentence is added to the dictionary, the variation of the artificial intelligence dialogue increases, and the user's desire can be further satisfied.

  Specialized semantic networks for products that implement a human body color numerical evaluation system are built on computer servers by customizing semantic networks by market, by target user, by service model, and by platform. Analyze with service model on computer server.

  The service model includes a navigation model, an entertainment model, and a self-mirroring model on a computer server, and the navigation model exhibits an information search function and a user-information matching function.

  The entertainment model can virtualize the personality of a virtual pet, character parameters (for example, whether to be aggressive or defensive), a breeding character, a talking partner, etc., as an application example to an amusement product.

  The self-mirroring model can confirm the personality, feelings, and preferences of herself (users) that she had not noticed before, and her newly discovered personality, emotions, and preferences that matched her personality, emotions, and preferences, talking partner, and secretary You can also guess.

  In the embodiment of the present invention, the user-preferred goods and services are estimated based on the information of the breeding character, the conversation partner, and the secretary on which the analysis results of the navigation model, the entertainment model, and the self-mirroring model are superimposed, and fed back to the platform. can do.

  The platform includes various products, and includes personal computers, personal digital assistants (PDAs), mobile phones, portable game machines, car navigation systems, next generation information terminals, and next generation home appliances. Since the user's personality, emotion, and preference can be confirmed according to the present embodiment, for example, the basic concept of an operation system for a next-generation home appliance network can be proposed (or output).

  The service model transmits provided information to the computer server of the service model by processing text and image information by the function of the editing unit of the computer server.

  The service model is a means for improving a product with a human body color evaluation system while drawing the user's preference through a color dictionary and providing it to any digital environment for the market.

  In the flow for determining the color information from the color chart, the color dictionary used in the human body color numerical evaluation system is configured to output the color information from among the color charts having the 1st to 127th color codes.

  The color chart has a white color code 64 at the center of the computer table, a red color code 101 at the back of the table, a blue color code 10 at the front of the table, and the front of the table. And a means for arranging a green color code 58 between the center and the center.

  In the blue group including 2nd, 3rd, 9th, 11th, 18th and 19th adjacent to 6 sides of the blue color code 10, 8 color codes belonging to the blue group are arranged and red In the red group including the 90th, 91st, 100th, 102nd, 110th, and 111th adjacent to the six sides of the color code 101, nine color codes belonging to the red group are arranged.

  Similarly, 10 color codes belonging to the green group are arranged in the green group including 46, 59, and 71 adjacent to the three sides of the green color code 58 and the white color code 64. White color codes adjacent to the six sides are also arranged in the range of 8 to 12 colors.

  In addition, the color chart has a color code that can be identified from red to purple, and can specify the color of words based on the movement and object of the human body and the human body and nature. For example, yellow color information is output from the words “say” and “giving”. Note that the same color information may be output even in the sound dictionary.

  The color chart can also specify the color from the relationship between the movement of the human body and the object or person. For example, a person who is a yellow person who emits irritating colors, outputs a vigorous color that encourages people and encourages people and things, and continues to work with an orange color that provokes creativity, outputs a patience color, and gives pleasure. It can be configured to output a color as described in (transmission).

  Similarly, from a balance with things and people, output a color that suits and gathers in a green color that emphasizes the person, care for the person in a blue color that serves and protects the person, and outputs a gentle color be able to.

  Furthermore, the color chart can also output color information that stimulates movements related to the touch of the human mouth and ears. For example, when a non-universal relationship arises from a person, a green color is heard to hear a third party's remarks, and a blue color that is less unilateral but does not affect the third person as much as yellow is output. Outputs a color that encourages the action of saying yellow with strong influence.

  Color dictionary processing includes classification processing for finally classifying words into groups of 8 to 12 colors from the color chart, word update processing for updating the word dictionary to reflect user preferences (color information), and artificial processing A system that executes color information transfer processing for passing color information to the intelligent dialogue engine and dialogue output processing for outputting dialogue from the artificial intelligence dialogue engine on a computer and returns to the main routine is targeted.

Correspondence between words in the sentence is determined by extracting red and blue colors from the color dictionary when the text contains the words “overhead” and “care”. You can adjust the parameters of categories such as physical, competitive, conductor, winner, courage to accomplish things, willpower, sexuality, vibrant, and entrepreneurs.

  The color dictionary executes code matching processing between the color code of the color number and the words of the text sentence, and can output a word or a fixed sentence belonging to red, and the red color of the first color information and the second color information Based on the blue color, green color information is output to the hue calculation result part.

  When the first color information is red, the second color information is red for red, yellow for orange, green for yellow, blue for green, indigo for blue, indigo for purple, and purple for purple To do.

  Similarly, based on the orange color of the first color information and the yellow color of the second color information, the orange color information is output. The green color is yellow, the blue color is green, the indigo color is indigo color, and the purple color is magenta. Output.

  Similarly, yellow color information is output based on yellow of the first color information and yellow of the second color information, and yellow is output in green, dark green in blue, blue in indigo, and red purple in purple. .

  Similarly, on the basis of the green color of the first color information and the green color of the second color information, the green and blue colors are dark green, the indigo color is indigo color, and the purple color is bluish purple.

  Similarly, on the basis of the blue color of the first color information and the blue color of the second color information, blue, indigo color is indigo, and purple is bluish purple, and the first color information is indigo color and second color. Purple color is output based on the purple color of the information, and purple color is output based on the purple color of the first color information and the purple color of the second color information.

  In addition, words that negative human impulses and actions negatively affect you and its surroundings fall from black to gray in the color dictionary and are painful (gray), kill (black), greedy (gray), forget (gray) ), Nuisance (gray), etc., and human trouble is treated as a psychological state corresponding to gray.

  For example, a device that performs emotion analysis by assigning a frequency band to a wavelength of color and quantifying it, for example, if it is speech, the user psychological state based on five senses information including visual angle, auditory sense, smell, touch and taste If it is classified as a yellow voice, it is cheerful, but if the frequency is too cheerful, it can be analyzed that the mood is too high and excited, and it can be applied to a device that can provide such a solution.

  On the other hand, words that indicate the smoothness of positive human behavior and the beauty of nature tend to be easy, quick, dawn, light, charm, and god are white. Also, the color classification varies from morning (white) to 3:00 (midnight) at night, and the color classification is linked depending on when it works most.

  If the user is a designer, purple / white color information for activities from midnight to dawn may be output from the color dictionary. For example, the color dictionary color-codes “battle” in red, “red is easy” in white, “win” in red and white, “defense” in green, and “green” in difficult indigo.

  It is also possible to construct a color word therapy by classifying small things into blue, blue and indigo with green as an important color, white not to forget indigo, and buds into colors.

  In the following embodiments, other elements can be referred to corresponding to the current dynamic color. For example, color information relating to the day of the week can be output from the color dictionary, Monday is white, Tuesday is red, Wednesday is purple, Thursday is light blue, Friday is green, Saturday is white, and Sunday is yellow.

  For words that are difficult to capture from meaningless kanji and meaning, it is possible to extract multiple color images from the sound table and determine the color, but the R color code is 238, G color code is 232, B color It is also possible to determine yellow from a plurality of color codes whose code is 117.

  You can also refer to words applied to the season. For example, in the spring, color information with a mixture of white as a pastel color is output, in summer, the color information is output with a cool sensation, in the autumn, warm color information is output, in winter, white or warm color information is output The human body color numerical evaluation system that can be configured on the computer and that runs on the computer inputs “violent” words or actions from the input unit, and the color chart number based on the color chart on the next page To decide.

  The color specified by the color chart number is decomposed into RGB information parameters and input to the color dictionary. The color dictionary outputs color information to the word dictionary based on the RGB information. A word dictionary can be grouped on the basis of RGB information, and patterns of the human body, emotion, and occupation corresponding to the word dictionary can be estimated.

  Analysis is performed based on the user's basic color (static color) using the color selected by the user from the word dictionary, the color used by the user, or a combination of colors. A comment corresponding to the pattern of the RGB word dictionary is extracted from the comment dictionary and inserted into the blank of the element text of the fixed pattern.

  Further, the human body color numerical evaluation system outputs an element text with a comment through an output unit. Element text with comments responds to words that claim color, such as text display, voice display, music display, and character display, specifies the color from the character of the word, and comments in the color system corresponding to the specified color An evaluation system for responding can be provided.

  In the word evaluation system, for example, words are input through a keyboard input, and the words are colored by a color dictionary.

  The color information selected in the coloring process is transmitted to the word dictionary, and a sentence corresponding to each color information registered in the word dictionary in advance is returned to the coloring process. For example, a pleasant or attractive sentence corresponding to the color information is returned to the coloring process.

  Further, the word dictionary transfers the color information transmitted from the coloring process to the numerical calculation unit. For example, the numerical value calculation unit adds the orange color information (100301) to the yellow color information (101002), divides it by the number of color information (two), and outputs the color information of the answer to the output dictionary configured in the database. Output.

  In the sentence coloring process shifted from the coloring process, the important words (words) included in the sentence are weighted. The processing shifts to the corresponding word selection processing, and the color information of the document corresponding to the weighted sentence is transferred to the output dictionary.

  Sentence pattern word embedding processing is connected to the output dictionary, and based on weighted color sentences and numerically calculated word color information, words corresponding to color information are embedded in the blank area of the fixed sentence pattern and completed A display process for displaying the sentence as text or voice is executed.

  The word evaluation system classifies words using a color dictionary that uses biological theory, determines the body energy state of the human body, and documents corresponding words that correspond to the body energy state (attentive / encourage / calm). Display processing. That is, it is intended for a system that digitizes emotions and psychological states corresponding to body energy states and outputs analysis results.

  The computer can organize the energy state of the human body according to color using a color dictionary, and can capture the user's actions from motion vectors. For example, a color from white to red indicates a state that appears in start-up and activity, and an orange color indicates a state that appears continuously.

  Similarly, yellow indicates a soft state that appears in joy, divergence, or giving, green indicates a state that appears in satisfaction or a balance of relationships with other things, and blue indicates a state of maintenance, stagnation, learning, or gentleness to protect. Indicates the state, indigo indicates an orderly state of teaching, applying, and purple indicates a hard, aged, and end of activity.

  Also, in sentences containing multiple words, say (yellow) include (green) indicates yellow over green, say (yellow) start (white red) indicates pastel green, and say (yellow) contend (red) ) Indicates orange, saying (yellow) matches (green) indicates yellowish green, saying (yellow) releasing (yellow) indicates yellow, saying (yellow) and (purple) indicates purple.

  Similarly, the word (yellow) finishes (purple) indicates purple, the word dating is exciting (yellow), reminiscent of the relationship (green) with people, yellowish green, and the word love is love. It means starting, and white, red, yellow from joy to purple pink.

  In addition, the color can be determined from the side of the kanji and the composition of the creation. For example, “food food” recalls appetite and shows orange. “Tree hen” reminds of planting, shows green, etc. The aura visualization system used for the human body color numerical evaluation system is equipped with a display that receives the electrical signal from the input section through the interface and displays the image processed image. It is intended for a system that branches the color information of the aura that is being used from the interface and transfers it to the color information input unit of the human body color numerical evaluation system.

  The aura color information described above is configured to be received after being branched from the interface. However, in the embodiment of the present invention, the configuration is not limited to this configuration. It is also intended for systems that are directly connected to and converted into electrical signals.

  The aura visualization system can measure a person's life data by placing the person's hand on the surface of the input unit. Human body aura chakra analysis is processed as follows.

  The aura visualization system can measure the life data of a person using a living body self-control sensor. This data analyzes the state of specific emotions and energy in real time.

  This aura chakra photo and report can express accurate bio data to the user in various visual graph formats. This biodata can be configured to be used for education and entertainment, and can also be used as an objective judgment material for the in-vivo energy state of the human body.

  The person's hand has two fingers arranged in the vertical direction with the little finger, the three fingers arranged in the vertical direction with the ring finger, the middle finger in the three electrodes arranged in the vertical direction with the electrode, and the index finger in the electrode A signal indicating the state of the internal organs of the internal organs is transmitted to the interface by bringing the thumb into contact with the electrodes provided on the right side of the input unit and the two electrodes arranged in the vertical direction to the three electrodes arranged in the vertical direction. To do.

  The aura color displayed on the display can reflect the psychological state, emotional state, personality, characteristics and characteristics of the human body.

  Dark red indicates physical, hard work, power, realistic, lively, calm, strength, stamina activity.

  Red indicates stimulation, physical, energetic, competitiveness, victory, self-control, sexual power, entrepreneurial activity, orange indicates productive, physical and creative expression, Shows activity of adventurers, business and fun.

  Dark yellow indicates scientific activity with analytical, intellectual, intentional detail, logical, and structured. Yellow indicates playful, cheerful, broad, creative, intellectual, pleasant, curious, and active. Green indicates the activity of social, nature, comfort, harmony, teacher, fondness and quick understanding.

  Catch green indicates fast understanding, goal-oriented, humorous, sensual, talk-loving, and the activity of the organizer. Blue color is kind, sensitive, loving, helping, faithful, peaceful, service minded, upbringing.

  The indigo color is clear, calm, deeply emotional, loving, faithful, spiritual values, and artistic activity. Violet colors indicate intuitive, artistic, sensual, theoretical, advanced, dreamy, charismatic and innovative activities.

  Light purple indicates activity of mediocre, mysterious, fantastic, sufted, sensitive, sometimes voting. White indicates extraordinary, transchromic, mediocre, peaceful, higher dimension, inspiring activity.

  In the enlarged image of the entire aura image on the display, the aura visualization system described above displayed the energy status of the six viscera and six limbs from the person's hand in color, but the entire aura image radiates from the upper body of the person using an imaging device not shown. The aura is displayed in color, and the strength of the aura and the color of the person are determined. The size of the aura can be determined by the distance between the person and the aura boundary.

  The aura of the entire aura image is displayed in the order of 6 colors of white, blue, green, yellow, orange, red from the bottom in the spectrum bar which is orange indicated by the arrow of the level gauge of the aura color spectrum displayed on the display. Color display.

  In addition, the spectrum bar displays not only the above-described six colors but also intermediate colors between the colors, and displays the state of the person. Since the person changes the emotion in response to the event, the level gauge of the aura color spectrum moves up and down in response to the change in the emotion.

  The real-time monitor displays a bar graph, and the current color changes to orange corresponding to the aura color spectrum. The height of the bar graph changes up and down in proportion to the aura boundary and the aura power. The illustrated bar graph has a height slightly exceeding level “80”.

  The display displays the aura power corresponding to at least 5 organs of a person in color and size on the seven bar graphs located in the upper stage. The activity of the six aura colors described above is visually represented.

  The leftmost bar graph displays red at level “100” and the stimulus activity is high, the right bar graph displays red below the level “100” and the stimulus activity is low, and the right bar graph displays yellow. The level “100” indicates that the activity level of playfulness is high, and the bar graph adjacent to the right indicates that the level of “100” is green, indicating that the social activity level is high.

  Similarly, the bar graph to the right of the bar graph displays blue at level “100” and has a high level of kindness, and the bar graph to the right displays blue at level “100” and has a high level of kindness and sensitivity. In addition, the bar graph on the right also displays blue at level “100”, and the activity of kindness and sensitivity is high, and the bar graph on the right side displays blue below level “100”, and the activity of kindness and sensitivity is slightly low. Is displayed.

  Next, first to thirteenth embodiments of the present invention will be described with reference to the drawings. In the following description of the drawings, the same or similar parts are denoted by the same or similar reference numerals.

  Further, the following first to thirteenth embodiments exemplify apparatuses and methods for embodying the technical idea of the present invention, and the technical idea of the present invention is the component parts. The material, shape, structure, arrangement, etc. are not specified below. The technical idea of the present invention can be variously modified within the scope of the claims.

(First embodiment)
As shown in FIG. 1, the emotion data providing apparatus according to the first embodiment of the present invention includes an operator terminal 3 for inputting operator answer information for user inquiry information, and a user terminal 2 through a first line 4a. A psychological analysis device 1 for receiving user inquiry information from the operator, receiving operator reply information from the operator terminal 3 through the second line 4b, and digitizing and analyzing the psychological state of the user and the operator.

  The first line 4 a transmits user data 5 to the psychological analysis device 1. The user data to be transmitted is exemplified as numerical data arranged on the time axis. However, the present invention can correspond to both an analog line and a digital line, and the coded numerical value can be displayed on the graph 7.

  The second line 4 b transmits the operator data 6 to the psychological analysis device 1. The transmitted operator data is exemplified as numerical data arranged on the time axis. However, the present invention can correspond to both an analog line and a digital line, and the coded numerical value can be displayed in the graph 8.

  The psychological analysis device 1 can generate a psychological mark 9 and display it on the operator terminal 3. Psychological mark 9 is selected based on psychological analysis results from Nikotama-kun series including No. 1 Hyokin mark, No. 2 refreshing mark, No. 3 serenity mark, No. 4 holding mark, No. 5 rage mark Is done.

  That is, the emotion analysis data generated by the psychological analysis device 1 can graphically display emotions including joy, anger, sadness, discouragement, and excitement of the telephone user and the operator.

  The operator data 6 or the graph 8 can also be used for the operator's mental care 10, but the present invention is not limited to mental care.

  The principle of the psychological analysis device 1 can be implemented based on psychological analysis by a voiceprint computer. The illustrated psychological analysis apparatus 1 can use a workstation or a personal computer, and the operator terminal 3 may be equipped with a psychological analysis engine.

  The voiceprint computer analyzes, for example, voiceprint data of former Iraqi President Hussein in order to pick up information hidden in the voice. Then, a graph is displayed on the display indicating that the frequency is plotted on the vertical axis, the time is plotted on the horizontal axis, the color is blue, the sound that the frequency has is stronger, and the frequency is weaker as yellow.

The voiceprint computer outputs, on a display on March 20, 2003, a graph in which “Amelie” in the line “America” in the speech by former President Iraqi Hussein is divided into voiceprints. In addition, a graph analyzing “Amelie” of “Americanian” in a voice-only statement broadcast by a television station on July 17, 2003 is also output on the display.
There are 10 matching points in the voiceprint data. There are also voiceprint matches for some of the words such as “Iraq” and “Jihad”.

  The Iraq War, which opened on March 20, 2003, was declared the end of the battle by President Bush (USA) on May 2, 2003. However, as of August 13, securing of the personality of former Hussein President of Iraq, which was the ultimate goal of the US military, has not been achieved.

  Former President Hussein is believed to have issued various statements while continuing his fugitive life. It is said that there are many shadow warriors in Hussein Yudai Taishu. Inevitably, the question will be whether the statement to be announced is by the former president of Hussein or someone has imitated it.

  On July 17, 2003, a Qatar television station broadcast the message of former President Hussein. This is only for audio without video. Is this “voice” really his / her own? The Japan Acoustical Research Institute, a private research institution known for voiceprint research, immediately analyzed this voice and said that this voice was the same as the voice of former President Hussein on March 20 (the first day of the Iraq War). I made a decision.

  In order to determine whether or not two voiceprints are the same, Japan Acoustics Laboratory normally checks 30 to 40 items such as frequency matching. Among them, if there is a match in 16 items, it is judged that it is the voice of the same person.

  Even if the voices of two identical twins (adults) are compared, the matching items are 8-9 items. If it is a voice of another person, there are at most two to three matching items. “Voiceprints have their own unique character, so voiceprints cannot be imitated,” says Dr. Matsumi Suzuki, director of the institute.

  “Three boxes” for creating voiceprints What is “voiceprint” in the first place? In a narrow sense, it is a graph that summarizes the voice of the voice, and in a broad sense, it is a collection of data obtained by analyzing the characteristics of the voice from various angles. The term “voiceprint” used by the Japan Acoustical Research Institute analyzes voice characteristics from various angles.

  Vocalization begins with the lungs pushing air. The air vibrates the vocal cords. What is created by the vibration of the vocal cords is a vibration sound of “sound” such as “boo”, which is not unusual. This vibration wave has no distinction between sounds such as “A”, “I”, and “U”.

  This vibration wave resonates with the “three boxes” from the vocal cords to the mouth, strengthening the vibration and resonating to form a “voice”. The “three boxes” are composed of a “vomit”, “oral cavity”, and “nasal cavity”. The pharynx is a connection from the vocal cords to the oral cavity and nasal cavity, the oral cavity is the space in the mouth, and the nasal cavity is the air in the nose.

  The oral cavity plays an important role in the “three boxes” to make words. This is because the volume of the air can be changed by the tongue. If the volume changes, the way the sound resonates (for example, how to resonate) also changes. By the way, there are words that do not vibrate the vocal cords, and that is called unvoiced sound. “P”, “t”, “k”, etc. in English phonetic symbols correspond to unvoiced sounds.

  On the other hand, words that vibrate the vocal cords are called voiced sounds. Its representatives are Japanese vowels such as “A”, “I”, “U”, “Engineer”, “O”.

  A combination of three frequencies produces a voice, and there are three bands called “formants (resonance frequency bands)” in the frequency of a voice uttered from a normal human mouth. It is a band in which the energy of the sound is strengthened by resonance, and can be defined as “first formant”, “second formant”, and “third formant” from the lowest frequency.

  Each formant has a certain width depending on the voice. For example, in the case of the voice “A”, the first formant is 600 to 800 Hz, the second formant is 1000 to 1400 Hz, and the third formant is 2700 to 3100 Hz.

  Some of the analysis requests received by the institute may include a recording tape that says "What is this talking about?" Many tapes are difficult to hear because of noise. For voice analysis, use formants and pray the position of the formants. For example, if the first formant is 700 Hz, the second formant is 1100 Hz, and the third formant is 2900 Hz, then you say “A” Analyzed.

  On the other hand, when using a computer to synthesize vibration waves with these three frequencies without using human voice, it is analyzed that the sound “a” was heard.

  Even cloned humans have different voiceprints. The formants that make up the voice have a range that is fixed for each sound, and the value that is taken depends on the person. Even with the same pronunciation “A”, the frequency value differs depending on the person, such that the first formant of the person A is 600 Hz and the first formant of the person B is 800 Hz. There are innumerable combination patterns of the first, second, and third formants even with the sound “a” alone.

  The cause of this difference arises from individual differences in the parts that make up the “three boxes” of the pharynx, oral cavity, and nasal cavity. Various factors are involved, such as mouth volume, lip thickness, and nose size. Dr. Suzuki argues that "the part that produces the voice is the person's personality that cannot be changed. This is the basis of the voiceprint."

  In addition, the voiceprint is also related to the speaker's own “kuse”. A habit such as intonation is created by the environment in which each person grew up. For example, the clone humanities appearing in (Science Fiction SF), etc., have the same size and shape of all parts of the pharyngeal oral cavity and nasal cavity. However, Dr. Suzuki insists that the same voiceprint cannot be obtained.

  In this way, no matter how much the figure is similar or the voice is similar, if the voice print analysis is performed, the voice of the person or the voice of the other person can be determined by the voice print computer.

  In Japan, voiceprints are also used as evidence in trials. The evidence value of the voiceprint is the second highest after the fingerprint and higher than the handwriting. In some cases, guilt or innocence is determined only by a voiceprint. In overseas cases. There is an example in which the Institute conducted a voiceprint appraisal of the assassination case of former Senator Aquino in the Philippines in 1983, overturning the announcement of the Philippine military administration at that time. As a result, distrust in the military administration increased in the Philippines, and it later developed into democratization.

  FIG. 2 is voiceprint data of the same person (20's female). The vertical axis represents the waveform (amplitude), and the horizontal axis represents time. Voice is represented by a series of repeated waveforms. In the twenties, the same waveform is taken in the first week and the second cycle.

  FIG. 3 shows voiceprint data of the same person (a 30th generation female). It differs from female voice print data (see FIG. 2) of women in their twenties in that a point (a box surrounded by a solid line in the figure) where the waveform cannot be maintained in the first week and the second week appears even in the same utterance.

  FIG. 4 shows voiceprint data of the same person (40th generation female). Even if the utterance is the same, it differs from female voiceprint data in the twenties in that a point (a box surrounded by a solid line in the figure) that cannot maintain a waveform appears in the first and second week periods. In particular, the tendency that the waveform cannot be maintained is strong in the 40's data.

  The voiceprint computer can estimate the height in steps of 5 cm. The voiceprint includes various intelligence. Information that cannot be grasped only by listening to the voice can be understood by analyzing it as a voiceprint.

  One of the information that can be read from the voiceprint is the height of the speaker. On the other hand, tall people tend to have low voices and short people tend to have high voices. Called “Phant's Law”, height and voice pitch (frequency of vocal cords) are inversely related.

  If you are tall, each part of your body will be bigger. Therefore, the vocal cords are also increased. Even if the air pressure sent from the lungs is the same, the taller person does not vibrate faster than the shorter person. This is connected to the bass. This is the same principle as taiko. Large drums produce low sounds, and small drums produce high sounds.

  By comparing with voice print data of many people in the past, height can be estimated as long as there is a voice. The accuracy is possible in steps of 5 centimeters with current technology. In addition, the chest circumference and waist circumference of the body do not affect the pitch of the voice as much as the height.

  The voiceprint computer can also include the technology for specifying the age in increments of 5 years. Age can also be estimated to some extent from the voiceprint. Human voice begins to deteriorate after age 25. For example, if you pronounce the word “A” for a long time, the waveform of the young person has less fluctuation, but this fluctuation tends to increase with age.

  The cause of the trembling is the muscle weakness. In order to utter a voice with a constant shaking “A”, air must be sent from the lungs to the oral cavity with a constant pressure, the vocal cord muscles must be kept constant, and the mouth shape must be maintained constant. As humans get older, their “fixedness” becomes difficult.

  In particular, the muscles that pull the vocal cords are weak and unstable, so that much energy is needed to keep them constant. As a result, the waveform appears to oscillate as you get older. Several years ago, “With the current split prayer technology, it is possible to specify the age in teenage units. In general, I would like to be able to specify the age in steps of 5 years old,” Dr. Suzuki said. Insisted.

  In addition, the voices of the elderly are often blurred. This is because the mucous membrane in the pharynx is uneven in the pharynx. When the air sent from the lungs passes through this irregularity, several vortices are formed. This vortex causes a faint voice.

  On the other hand, except for the components that create such shaking, it is possible to reproduce the voice of an elderly person on a voiceprint computer. The words can also be determined by physical condition. The physical condition of the speaker is one of the information that appears in the voiceprint.

  First, when inflammation occurs in the pharynx, the same phenomenon as in the elderly occurs. In other words, unevenness occurs in the mucous membrane due to inflammation, and a vortex is generated, resulting in a faint voice. In addition, when the body is in poor condition, the abdominal muscles do not get moistened, and a large amount of energy is used for voicing using the abdominal muscles. As a result, when the voice is uttered without using the abdominal muscles, the frequency component of 2000 to 3000 Hz becomes extremely weak when the voice is subjected to frequency analysis.

  Among the voices analyzed by the institute, a good example of a voiceprint representing changes in physical condition is voice data of Bin Laden, the leader of al-Qaeda, an international terrorist organization.

  When Bin Laden's statement, which has been searched by the US military since the September 11, 2001 terrorist attacks in the United States, is analyzed by voiceprints, there is a weak frequency component of 2000-3000 Hz. For this reason, the institute breaks down that Mr. Bin Laden was in a state of health while continuing his escape life.

  Voiceprint analysis begins with searching for the same voice. How the voiceprint is broken up is case-by-case. When a voice print appraisal request is received for a tape containing the voice of a person from the same laboratory, the same word is extracted from the voice of the same person recorded in the past. Especially when it is not in Japanese, it takes time to find the same word. Without the same word, the accuracy of voiceprint evaluation may be significantly reduced.

  Therefore, paying attention to individual habits, search for the same voice. For example, when you introduce yourself, you extract the habit of “Eh” from a person named “Eh, I am I.” The time required for voiceprint analysis is sufficient if it is 1 to 2 seconds. Once the word is extracted, noise is removed. Also, the voiceprint computer has different frequency characteristics depending on the recorded equipment. There are frequency characteristics unique to each of telephone, broadcast waves, and direct recording on a recording tape. Perform various corrections and perform analysis.

  The voiceprint computer stores voice data broadcast on October 8, 2001 and voice data acquired on December 14, 2001 as voiceprint data suggesting Mr. Bin Laden's poor physical condition. The frequency is plotted on the vertical axis and time is plotted on the horizontal axis. The bluer the waveform color, the stronger the sound with that frequency and the bluer the weaker the sound with that frequency.

  Although it can be confirmed that both are voice print data of Mr. Bin Laden by analysis from other angles, when the voice data acquired on December 14, 2001 is analyzed, the frequency around 2000 Hz to 3000 Hz is extremely weak. This is because the frequency in the vicinity of 2000 Hz to 3000 Hz is weakened when the user feels sick and speaks without using the abdominal muscles.

  The early speech synthesizer produced by Dr. Suzuki is wooden and has a hollow structure, with the upper two round holes imitating the nostrils and the lower square hole imitating the mouth. Words can be created by sending vibration waves from the bottom of the box and skewing the beat trees on both sides.

  A voiceprint computer displays voice data generated by voice synthesis as a waveform on a display. For example, monalisa voiceprint data (the vertical axis is frequency and the horizontal axis is time) is displayed. This is audio data that was restored from the Mona Lisa's skull from the painting “Mona Lisa” by Da Vinci and pronounced “Sono Elizabetta Monnalisa” (I am Elizabetta, who is called Mona Lisa).

  The voiceprint computer can also output voiceprint data of the Cretaceous dinosaur Parasaurolophus with Tosaka. In the case of Balasaurolophus, two types of voices are synthesized according to the skeleton. A low voice is considered a male call and a high voice is considered a female call, but now they are not living on the earth and cannot be confirmed.

  On the other hand, Dr. Suzuki argues, “The tape recorded on site contains a lot of information in the back noise”. In the case of a tape recorded indoors, the voiceprint computer analyzes the back noise and can determine from the reverberant sound whether there is a floor material, the presence or absence of air-conditioning equipment, and a picture in the recorded background. The voiceprint analysis can understand the number of people in the same room.

  The voiceprint computer also has various times for analysis, and outputs the results at several times in the earliest. In addition, Kenji Miyazawa's voice can be restored using speech synthesis technology. Currently, the mainstream of car navigation systems and station announcements is based on recording and synthesis technology.

  This is a technology that actually records human voices and connects the recorded voices to create dialogue. For example, in car navigation systems, etc., names of intersections across the country are recorded in advance and combined to create dialogue. This is a technology made possible by the development of large-capacity recording media such as DVD-ROMs and hard disks. However, it is necessary to record a human voice as a precondition, and it is impossible to make an abandonment that has not been recorded.

  By reproducing the “three boxes” of the pharynx, oral cavity, and nasal cavity, it is possible to create a vibration wave that is the same as the frequency of the vocal cords and create an artificial voice that does not depend on the recorded voice. A technology that focuses on this point is a speech synthesis technology called “voice vocal tract simulation” developed by the institute.

  The principle of speech synthesis technology is simple. The path of the voice, that is, the pharynx, oral cavity, and nasal cavity that make up the “vocal tract” can be created as a model, and vibration waves having the same frequency as when the vocal cords are vibrated are sent from the back of the pharynx.

  After that, you can make a voice by changing the volume of the oral cavity. Of course, in this case, it is not possible to include the “quick” of the person's way of speaking, but the so-called “voice without inflection as the robot is speaking” is created based on the speech synthesis theory.

  The laboratory originally reproduced the vocal tract with a wooden model to verify this speech synthesis technology. As a result, it is possible to reproduce the sound. "Even a model of a tree can make a voice. Actually, I thought that a voice wasn't made by a great mechanism. I think that if I make this, I can synthesize voice better, ”recalls Dr. Suzuki.

  What is needed to reproduce a voice with an electric circuit is the skeleton data of “three boxes”. Even the skeleton data can restore the voices of greats and celebrities that do not exist. To reproduce the voices of past greats and celebrities, first calculate skeleton data from bronze statues and photographs.

  Based on the skeleton data, “three boxes” should be reproduced and synthesized with a voiceprint computer. The voices reproduced by Tsukasa Research Institute include recent movie actor Yujiro Ishihara (1934-1987) and poet / fairy tale writer Kenji Miyazawa (1896-1933). It is a finish that we recognize. In addition, the skeleton is inferred from the painting “Mona Lisa” by Leonardo da Vinci (1452-1519), and the voice of a woman who is said to be a model of “Mona Lisa” is restored.

  Voiceprint computers can create a standard for dinosaur voices. The institute collaborated with dinosaur researchers Philip Curly (Royal Tyrell Museum, Canada), Dr. Martyn Lockley (University of Colorado, USA), Dr. Masaki Matsukawa (Tokyo Gakugei University), etc. Is reproduced.

Currently, the standard of dinosaur voices we hear in movies is based on research by the Institute.
If the subject is a dinosaur, it begins by analyzing the skull fossils with a CT scan. By performing a CT scan, the volume of the cavity in the skull can be calculated. In other words, you can get the data of “three boxes”.

  Dr. Suzuki assumed that the dinosaur had no vocal cords, referring to a dinosaur researcher's point. Based on the hypothesis that the dinosaur was vibrating the pharyngeal mucosa as an alternative to the vocal cords, it creates a vibration wave and reproduces the voice of the dinosaur on a voiceprint computer.

  The institute reproduced the voices of Tyrannosaurus Rex, Stegosaurus and Parasaurolophus. Dr. Suzuki talks about Parasaurolophus in particular. "Dinosaur researchers have long hypothesized that Parasaurolophus with its characteristic scorpion was a horn-like voice. The voice synthesized at this institute was also like a horn. The opinions of the dinosaur researchers and the acoustic researchers agreed. "

  For voiceprint computers, the development of speech recognition technology is the next issue. The Institute conducts all kinds of research related to sound in addition to speech. In recent years, at the request of JASRAC (Japan Music Copyright Association), he participated in an international digital watermark development project and examined watermark technology. Digital watermarking is a technology that embeds a signal that prevents copying in digital content data such as music.

  We asked Dr. Suzuki about future issues in acoustic research. “Now, the most desirable is the development of speech recognition technology. It is a technology that allows computers to recognize words accurately while judging the context before and after, just like human questions. It is used in limited fields such as tourist information, but it is not yet universalized. "

  The Institute's activities, which began with voiceprint analysis, led to speech synthesis technology that gave machines a mouth. If new speech recognition technology is completed, the machine will have ears. Then, there may come a day when you can interact with the machine as if you were talking to a human. Future research of the Institute is expected to continue the challenge of “sound”.

  Dr. Matsumi Suzuki, director of the Japan Acoustical Research Institute, says, “It is important not to lose curiosity like a child forever”.

  The character-type conversation system used in the following embodiments is a system for exchanging character-like conversations with a user using a computer. For example, a humanoid robot, a pet-type robot such as an animal, and any other form of robot set to target.

  As shown in FIG. 5, the character-type conversation system exemplified as the emotion data providing apparatus according to the first embodiment of the present invention includes a semantic network database 94 containing a large number of vocabularies and an SDD module for generating free sentences. 97, a fixed topic module 91 that develops a conversation centering on a specific topic, and a conversation topic coherency between the SDD module 97 and the fixed topic module 91 is maintained and managed, and the conversation can be continuously expanded. The first and second connection modules 93 and 99, the user information management module 89, and the character management module 90 are provided.

  The character-type conversation system provides emotion data by a character-type conversation system having a first continuation check unit 92, a second continuation check unit 98, a related word / link word check unit 95, and an SDD function check unit 96.

  In the semantic network database 94, vocabulary of about 5000 words is registered together with pronunciation / grammar information, and includes a noun table and a predicate table. Field categories provided in the noun table include target nouns, link nouns, superordinate concepts, subordinate concepts, associative vocabulary, predicates, part of speech information, attribute information, and the like. The field category provided in the predicate table includes a target predicate, a predicate noun, an antonym predicate, a link noun, part of speech information, attribute information, and the like.

  This semantic network database 94 plays an extremely important role in executing a plurality of modules. That is, in order to realize topic development / return / development and generation of coherent sentences, it is necessary to impose specific syntactic constraints, but the semantic network database 94 itself maintains a certain degree of cohesion between vocabularies. Therefore, it is an optimal knowledge management database for realizing the purpose of this character type conversation system.

  The SDD (Speaker Driven Discourse) module 97 realizes the function of generating free sentences according to the user's response as well as providing various topics spontaneously in order to ensure the autonomy of the character type conversation system. Therefore, the development pattern information of the discourse activity performed by humans is stored, and an optimum pattern is selected according to the situation in the discourse, and the conversation development is performed based on the pattern.

That is, the SDD module 97 maximizes each feature in the Japanese discourse grammar and realizes an autonomous discourse function with a minimum discourse constraint. In Japanese discourse grammar, there is a characteristic restriction that the known information is omitted and cannot be expressed in the sentence. In other words, once the topic and the listener are determined, it is not necessary to express the listener's name and topic. For example, A: Did you go to (your) company yesterday?
B: (I) went to (my company).
A: Are you tired?
B: Okay.
A: Did you work overtime?
B: No.

  In the Japanese discourse grammar from the above-mentioned conversation, once “who is the topic” is determined, it is not necessary to specify the person as a pronoun like English. In addition, once a topic is determined (for example, “company”), other information (for example, “overtime”) is determined in relation to the topic frame (knowledge framework) as much as possible.

  Therefore, when A asks "What is overtime?", B makes an utterance based on the judgment that "I should have asked if I worked overtime at my company." In other words, there is a mechanism in Japanese where topics and participants are decided, and when certain shared knowledge is established, missing information parts that do not appear in words are supplemented and understood by each participant. .

  The SDD module 97 basically applies this mechanism to perform conversation development. In the discourse progress process, a new development part in which a new topic and a new participant are recognized and a conversation based on the existing information. To make the system (robot) comply with the conditions required for each process of new development and existing information development, to realize "approximate" natural conversation. In other words, as long as the minimum required discourse conditions are observed in each process, a mechanism is applied in which the user unconsciously supplements and understands missing information and related information.

  By the way, human discourse does not always change while maintaining a certain level of mobility (speed), but in the exchange of specific topics, the mobility is rather lowered, and the information area tends to narrow and the information depth increases. . That is, the sentence generation dynamics and the information depth are inversely related.

  Among them, the fixed topic module 91 is provided to secure the information depth while ensuring the mobility by the SDD module 97. Therefore, the fixed topic module 91 has various information depths set for a plurality of individual topics (over 40), and a plurality of response sentences are prepared at each depth. Then, by combining the SDD module 97 and the standard topic module 91, it is possible to simultaneously secure conflicting discourse mobility and information depth.

  In the first and second connection modules 93 and 99, the discourse cohesiveness maintained and managed in order to enable continuous development of conversations is syntactic / semantic cohesiveness in the same sentence in the order of narrowing the area. Second, cohesiveness between user utterances and robot response sentences, third, cohesion between topics and spoken sentences in discourse, and fourth, cohesion between known user information and robot on-site utterances It can be classified into four levels of sex.

  Of these, the syntactic / semantic cohesiveness within the same sentence is primarily the sentence format that the robot generates freely. In this character type conversation system, syntactic cohesion is ensured by causing the SDD module 97 to generate a sentence centered on the main predicate syntax based on the analysis of the Japanese syntax.

  As a result, vocabularies stored on the semantic network database 94 while maintaining cohesiveness can be reflected in the sentence without destroying their semantic cohesiveness. Further, for a predicate that requires the subject noun to have a feature of [+ human], a flag of [+ human] is provided in the database to narrow down co-occurrence relationships in the main description.

  Secondly, the cohesiveness in the user utterance sentence and the robot response sentence is realized by hedge management. In other words, “hedging” is an action that is spoken as “connecting” or “pillow” before speaking with substantial content, and plays a major role in maintaining discourse cohesion.

  That is, instead of making the robot speak immediately after the user's response sentence, put a phrase that confirms, empathizes, or refutes the attitude to the user's proposition, and then continues the utterance sentence, so that the level difference between sentences To ensure unity.

  Therefore, in this character type conversation system, an appropriate response is returned according to the user's response in the SDD module 97, in the fixed topic module 91, in the first and second connection modules 93 and 99, and between them. Hedge management.

  Third, the cohesiveness between the topic in the discourse and each sentence uttered is realized by the concept of topic management. That is, in this character type conversation system, when there is a preceding topic, the appearance of a new information vocabulary that overtakes it is limited.

  Fourth, regarding the cohesiveness between known user information and on-site utterances of the robot, the user information learning function that stores and maintains the content, utterance attitude, and information that the user has spoken before as a log, and the robot utters This is realized by providing a generated sentence management log function for storing and maintaining the information. That is, in this character type conversation system, the user information management module 89 has a user information learning function, and the character management module 90 has a generated sentence management log function. Knowledge established with the robot is reflected in later conversations.

  This character-type conversation system has two types of functions for character formation, one is performed by dynamic expression of a specific character kicked by a keyword, and the other is stored. It is performed by expressing a specific character based on the static data of the user response.

  The dynamic expression of a specific character becomes a keyword when the user utters a specific word, and a character corresponding to the type of the user utterance is realized in the robot utterance. For example, “gag fire” and “return of nose” can be mentioned.

  The expression of a specific character is a function in which a specific character is expressed at a certain frequency when the knowledge managed by the user information learning function satisfies a certain condition. This is mainly performed by the character mode management function.

  For example, if a user frequently uses a word corresponding to the “father word group”, the robot easily enters the “father mode” at a high frequency.

  On the other hand, a dynamic character is also expressed by the time control function. For example, when a time schedule set by the user such as a wake-up time is ignored, an utterance with a specific emotional color is selected.

  These character characteristics are effective in strengthening the impression given to the user by being synchronized with the facial expression displayed on the display of the robot face. When the character-type conversation system is configured as a robot of an arbitrary form such as a human-type or a pet-type, an LCD device for expressing the facial expression of the robot to the user and an LED device for expressing the emotion of the robot to the user are used. It is preferable.

  The operation of the robot conversation system exemplified as the emotion data providing apparatus according to the first embodiment will be described with reference to FIGS. In the robot conversation system, in the first utterance step, first, the SDD module 97 is prepared with a confirmation type sentence list, a report type sentence list, a question type sentence list, an evolution type sentence list, a conversion type sentence list, and a situation setting type. From the sentence list, first, the question type sentence list is selected, and the sentence type “$ user name $ Nittotte N1 (noun) tteudouimonokana” is further selected (step SR1). .

  Next, the SDD module 97 quotes “Suzuki” from the user information management module 89, embeds “Suzuki” in the sentence-type user name (step SR2), and quotes “Sake” from the semantic network database 94. Then, “sake” is embedded in the sentence pattern N1 (noun) (step SR3), thereby generating a sentence “What is sake for Suzuki?” (Step SR4) and causing the robot to speak.

  At this time, the SDD module 97 refers to the request information type table (step SR5), thereby recognizing that the uttered sentence is a type of sentence requesting information for filling the Wh information.

  When the user responds, for example, “I don't know” to the first utterance (step SU1), the SDD module 97 determines that the second connection module 99 hits the Wh information based on the above-described request information type recognition. Refer to the word list table (step SR11), check the user response hit word (step SR12), refer to the user reaction interpretation table (step SR13), and the user's first utterance "I don't know" is "Unintelligible" As a message. Then, the corresponding hedge is checked (step SR14), and as the first hedge utterance, for example, the robot outputs the hedge output “Jasa”.

  Subsequently, the SDD module 97 refers to the second step connection table (step SR21), selects the sentence type “V1 (verb) (na) notte | V2 (verb) kotonano cana”, and “sake” from the semantic network database 94. The words “drink” and “fun” having semantic cohesiveness are quoted and embedded in V1 and V2 (verbs) respectively (step SR22), and a sentence “Is drinking fun?” Is generated ( In step SR23), the robot speaks.

  The SDD module 97 responds to the second utterance, for example, when the user responds “Yeah, it's fun” (step SU11), the continuation check unit has related words and related topics in the standard topic module 91. (Step SR31), and if it exists, the process branches to a continuous flow.

  On the other hand, if the related word or related topic does not exist in the fixed topic module 91, the semantic network database 94 is referred to (step SR32), and whether or not the related word or related topic exists in the fixed topic module 91 or the SDD module 97. To check.

  Then, the related word / link word check unit 95 performs a related word link check (step SR33), and if a related word exists, the process branches to either the fixed topic module 91 or the SDD module 97. At that time, the related word / link word check unit 95 determines whether or not to bypass the first or second connection module 93, 99 according to the consistency between the utterance sentence pattern and the branch destination in the second utterance step (step). SR34).

  Subsequently, the SDD function check unit 96 performs an SDD function check (step SR35), and if there is a related word or related topic, the process branches to either the fixed topic module 91 or the SDD module 97 (step SR36). . On the other hand, if there is no related word or related topic, a new topic is randomly started (step SR37).

  FIG. 9 is a table showing the contents of the first and second connection modules 93 and 99 in the SDD module 97, and FIG. 10 is a table showing the hedge connection in the SDD module 97. In FIG. Means negative, P means positive.

  As shown in FIG. 10, patterns 1 to 6 correspond to a confirmation type sentence list and a situation setting type sentence list. That is, when the first utterance of the robot is a confirmation-type sentence, the pattern 1 (see only numbers) that can become the second utterance of the robot is selected by looking at the pattern 1.

  Similarly, if the first utterance of the robot is a report-type sentence, the pattern 2 is seen. If the first utterance of the robot is a question-type sentence, the pattern 3 is seen. Further, the first utterance of the robot is developed. See pattern 4 for a type sentence, see pattern 5 if the first utterance of the robot is a conversion type sentence, and see pattern 6 if the first utterance of the robot is a situation setting type sentence. Then, each sentence pattern (only numbers are shown) that can be the second utterance of the robot is selected.

  The character-type conversation system applied to the first embodiment is a character-type conversation robot, that is, speech recognition, word spot function, artificial intelligence, automatic sentence generation function, single-tone recognition, memory function, remember words It can be implemented as a communication robot or communication application means capable of chatting freely, and becomes a multipurpose communication application such as an information home appliance (for example, an Internet refrigerator or an Internet recorder) equipped with the communication application means or an interior.

  For example, when the user talks to the communication robot, “Today is good weather,” the words “Today”, “Good” and “Weather” are recognized by the word spot function of voice recognition. From that information, artificial intelligence searches related words and returns a reply using the automatic sentence generation function.

  In addition, a communication robot equipped with a communication application completes a flow of conversation with a sense of incongruity, such as “It was raining yesterday.” The word spot function uses not only one word but also two or more words as recognition words, and artificial intelligence picks up related words from the previous conversation. You can expect a reply.

  Furthermore, by speaking to each user's unique information, it is possible to store it properly and reflect it in the subsequent conversation, and to realize a live conversation with the user. For example, in order to reflect the memorized memory in the conversation like a human being, it may be said that “I was talking about this last year” in a year later.

  Since the communication robot can memorize not only words but also a schedule, it can have an appropriate conversation in a daily flow, a weekly flow, or a yearly flow. For example, the service is provided by singing “It's morning, get up”, “Today is the day to throw out trash”, or “Happy Birthday (user name)”.

  In addition, by lip-synchronizing more than two frames for each character, the conversation partner can be changed while changing abundantly according to the content and language of the conversation.

  In the first embodiment, those skilled in the art can fully understand that the communication robot as the emotion data providing device further evolves due to the expandability by the ROM cassette and the connection with the personal computer or mobile phone using the serial port. Will.

  Furthermore, in addition to humanoid robots as emotion data providing devices, pet-type robots such as animals, and other arbitrary robots, for example, personal computers and portable information terminals can be targeted as character-type conversation systems, As an example of an application example of a mobile terminal where two or more speakers are conversing and grasping the psychology of the communication application and intervening, the mobile terminal is angry by the “Gamigami boss” and hangs up. However, as if he had a personality, he was able to enter into the arbitration between two fighters, or to intervene in support of the lover business.

(Second Embodiment)
The character-type conversation system used in the second embodiment of the present invention is provided with an advertisement scenario sub-module that executes a predetermined advertisement scenario according to a predetermined trigger in the fixed topic module 91.

  As shown in FIG. 6, the advertisement scenario submodule exemplified as the emotion data providing apparatus according to the second embodiment of the present invention includes an advertisement scenario database 156, a user response interpretation table 157, and an advertisement appearance frequency pattern table 155. And an advertisement execution trigger generation unit 151 and an advertisement scenario processing unit 150.

  The advertisement execution trigger generation unit 151 includes a first trigger generation unit 152 that generates an advertisement execution trigger at the time of a topic change after a predetermined time has elapsed since the system startup, and an advertisement when a predetermined process such as a wake-up process occurs. A second trigger generation unit 153 that generates an execution trigger, and a third trigger generation unit 154 that generates an advertisement execution trigger based on an advertisement appearance status based on the advertisement appearance frequency pattern with reference to the advertisement appearance frequency pattern table 155 I have.

  The advertisement scenario processing unit 150 executes the advertisement scenario stored in the advertisement scenario database 156 in response to an advertisement execution trigger generated from the first trigger generation unit 152, the second trigger generation unit 153, or the third trigger generation unit 154. To do. This advertising scenario is a scenario of a series of interactions between a robot's utterance including the utterance of a predetermined advertising topic and a user's reply, no matter how the conversation progresses due to the user's reply, It is intended for a system in which a robot can speak a message including a predetermined advertisement topic in a natural conversation flow.

  The advertisement scenario sub-module includes a user attribute database 158 for referring to user attributes when executing the advertisement scenario, and the advertisement scenario processing unit 150 interprets the user's reply accompanying the execution of the advertisement scenario. The user information is stored in the user attribute database 158 by analyzing and acquiring the user's emotion and psychological state regarding the predetermined advertisement topic by collating with the table 157.

  The advertisement scenario sub-module acquires in advance without executing the advertisement scenario, based on the user information acquired through the execution of the advertisement scenario (such as the user's emotion and psychological state regarding the advertisement topic, preferences, preferences, hobbies, etc.). The user information can be reflected in the execution of the advertisement scenario as user attribute information.

Further, the advertisement scenario sub-module, when a vocabulary inappropriate for execution of the advertisement scenario is included in the words stored in the noun table or the predicate table of the semantic network database 94, when setting the advertisement scenario, An emotion data providing apparatus for deleting / correcting inappropriate vocabulary in the semantic network database 94 is provided.
With reference to FIG.6 and FIG.11, operation | movement of the advertisement scenario submodule illustrated as an emotion data provision apparatus which concerns on the 2nd Embodiment of this invention is demonstrated.

  When the conversation robot is interrupted for a specific time after the system is started (step S100) (step S105), the conversation robot performs the following utterance (that is, forced utterance at the time of topic change or specific process occurrence). .

(A) At the time of topic change after a predetermined time has elapsed since the conversation robot was activated,
R: “Oh, I talked too much and I was so cute, XX, let ’s drink cola.”

(B) When the wake-up process occurs in the conversation robot,
R: "Good morning, it's time to wake up, don't fall asleep forever, and drink with cola."

  On the other hand, when the conversation robot proceeds to the fixed form / SDD branch (step S101), the topic T is first determined (step S111). Such topic determination is performed when the conversation robot advances the fixed / SDD branch (step S101) to a fixed form in a predetermined process of the character-type conversation system, or by topic change by the user (step S110). Is also done.

  Next, the conversation robot refers to the advertisement appearance frequency pattern table 155 (step S112), and when an advertisement execution trigger occurs, searches the advertisement scenario database 156 for the topic T (for example, “drink”) (step S113). ), A predetermined advertisement scenario for extracting the utterance of the advertisement topic A (for example, “cola”) from the topic T is determined (step S114).

  When the user attribute is unnecessary (“NO” in step S120), the conversation robot does not reflect the user utterance for executing the advertisement scenario (that is, user-driven or random user attribute selection when selecting a specific topic) (Step S121).

(C) When the conversation robot selects a drink topic,
R: “What kind of drink do you like XX?”
U: “Cola”
R: “After all, I love cola! That feeling of swoosh is irresistible.”
U: “That ’s right.” R: “Shwaswah.

(D) When the conversation robot selects a drink topic,
R: “What kind of drink do you like XX?”
U: “(other than cola)”
R: “Hey, what is that, my favorite is cola after all, XX do you like cola?”
U: “Separate…”
R: “What? That Schwashwa, if you imagined, I wanted to drink right now, please buy me a cola!”

  In this way, the conversation robot naturally utters a predetermined advertisement message (for example, “Please buy cola”) including the advertisement topic A (for example, “Cola”) regardless of what kind of reply is returned from the user. It is possible to guide in the direction to do.

  When the conversation robot needs user attributes (“YES” in step S120), the conversation robot refers to the user attribute database 158 (step S131) and performs the following conversation for executing the advertisement scenario (ie, in advance). A conversation that reflects the acquired user attributes is performed (step S132).

(E) If the user is a user who likes beer,
R: “Yes, did you like beer?
U: "Sooda"
R: “Since I ’ve been drinking too much recently, it ’s better to occasionally drink cola.”
U: “Why?”
R: “Because I'm doing something like that, Sooda, OO, instead of drinking beer from now on, if you drink cola, it'll be good for your health.”

(F) In the case of a dieting user:
R: “I'm sure you were on a diet, XX?”
U: "Sooda"
R: “After all, do you also get sweet drinks?”
U: “Well,”
R: “That's fine with diet cola, it ’s low in calories.”
U: “But ~”
R: “Don't overdo it for a diet, do n’t drink diet cola because you say I ’m fine.”

(G) In the case of a user who likes cola,
R: “Yes, did you like cola, XX?”
U: "Sooda"
R: "If so, do you collect extra figures?"
U: "I'm collecting"
R: “Once collected, I will not stop until it is completed”
U: "Sososo"
R: “Well, you can collect figures and it's delicious to drink, and cola is the best drink for XX!”

  When the conversation robot ends, the conversation robot refers to the advertisement appearance frequency pattern table 155 (step S141). When an advertisement execution trigger occurs, that is, when the current topic is continued for advertisement expression ( “YES” in step S142), the process returns to step S114, and a new advertisement scenario that draws the utterance of the advertisement topic A (eg, “cola”) from the current topic T (eg, “drink”) is determined, and the conversation is executed. . Further, when the conversation robot does not continue the current topic (“NO” in step S142), the conversation robot checks whether or not to continue the loop within this topic (step S143), and when it continues the loop (“ YES ”), the process returns to step S114. On the other hand, if the loop is to be ended (“ NO ”in step S143), the process branches to the end of the immediately preceding topic (step S103).

  Further, deletion / correction in the case where the semantic network database 94 includes vocabulary inappropriate for execution of the advertisement scenario is performed as follows. That is, when the system (robot) tries to speak automatically by using the SDD module 97, a link such as noun → predicate (verb, adjective, etc.) related to the noun exists roughly in the semantic network.

  For example, if the noun “Cola (juice)” is linked to the predicate “Drink”, “Sweet”, or “Fat”, the conversation robot generates the sentence “Cola is fat”. there is a possibility. Therefore, in order to avoid this, the predicate link “fat” is deleted and a predicate such as “fresh” is added, reflecting the opinion of the advertiser.

  In addition, by using an LCD device that can draw and display the facial expression of the conversation robot to the user, company logos, product images, etc. can be displayed instead of the robot's facial expression according to the utterance content of the system. it can.

  The conversation robot can acquire various information such as a user's emotion, psychological state, preference, preference, and hobby regarding a predetermined advertisement topic by using the user information acquisition method, which can be used for a questionnaire survey.

When the conversation robot performs user information acquisition processing in the YES / NO format,
R: “Do you like cola, OO?”
U: “Yes”, “Yes”, “Like”, etc. In this case, it is recognized as YES.
U: “I dislike”, “No”, “None”, etc. In this case, it is recognized as NO.
U: If the reply does not fall under either category, it is recognized as neither, and can be removed.

In addition, when the conversation robot performs user information acquisition processing in the option format (multiple answers are not possible), R: “XX is a drink from Coca-Cola, whichever you like, cola, fanta, sprite… Or R: “What do you like when XX is a Coca-Cola drink?”
U: “Cola”, “Coca-Cola”, etc. In this case, it is recognized as a cola.
U: “Fanta” etc. In this case, it is recognized as a Fanta.
Thus, if the options are registered in advance, the answer can be recognized by the user's normal utterance. In addition, in the case of a user reply that is not in the assumed option, a character string is input by single-tone recognition (similar to the [other fields] of the questionnaire using characters).

Furthermore, when the conversation robot performs user information acquisition processing in the option format (when multiple answers are possible / up to two can be selected),
R: “Why XX likes cola?”
U: “It's delicious and refreshing” …… In this case, it is recognized as “delicious” and “fresh” by the word spot.
U: “It ’s delicious…” …… In this case, it is first recognized as “delicious”, and R: “Is there any other reason besides delicious?”

  On the other hand, if the answer is YES, an additional reason is recognized, and if the answer is NO, only “delicious” is set as the answer. Handling of options and processing in the case of a user reply that does not exist in the assumed options are the same as in the case where multiple answers are not possible.

In the case of user information acquisition processing in the user option format (multiple answers possible / 3 or more selectable)
R: "Who is your favorite actor? Tell me for the time being."
U: "Actor A"
In this case, “Actor A” is the first answer, and R: “Are there other actors besides actor A?”
On the other hand, if YES, it recognizes additional actors and loops to the MAX value. If NO, answer “Actor A” only. Handling of options and processing in the case of a user reply that does not exist in the assumed options are the same as in the case where multiple answers are not possible.

  When the user is in a free answer format, it is possible to respond by inputting characters to the conversation robot one character at a time using single-tone recognition.

  The character type conversation system exemplified as the emotion data providing apparatus according to the second embodiment is equipped with a character type conversation robot, that is, voice recognition, word spot function, artificial intelligence, automatic sentence generation function, single sound recognition, and storage function. It can be put into practical use as a communication robot that can learn words and talk freely.

  For example, a character-type conversation robot analyzes the user's emotions and speaks "Today is a good weather." When using the word spot function of voice recognition, the words "Today", "Good", and "Weather" are key words. Recognize From that information, artificial intelligence searches related words and returns a reply using the automatic sentence generation function. “It was raining yesterday,” “I feel good when the weather is good,” and so on. The word spot function uses not only one word but also two or more words as recognition words, and artificial intelligence picks up related words from the previous conversation. You can expect a reply.

  In addition, the character-type conversation robot analyzes the user's emotions and lip-synchronizes two or more frames for each character.

(Third embodiment)
As shown in FIG. 12, the emotion data providing apparatus according to the third embodiment of the present invention shows the user's psychological state based on five senses information including vision 12, hearing 13, smell 14, touch 15, and taste 16. When a numerical value that is psychologically depressed by the user 11 is detected based on the user history information stored in the user information management module after being digitized and analyzed, the user 11 outputs an encouragement message using characters, pictures, or sounds. When psychologically uplifting, a psychoanalysis device that outputs a dream recall message is provided.

  The psychological analysis device may detect the five senses using a sensor and input information on the five senses.

(Fourth embodiment)
As shown in FIG. 13, the emotion data providing apparatus according to the fourth embodiment of the present invention includes an operator terminal 3 for inputting reply information of operators 17a and 17b for inquiry information from a user through a line, and a user terminal 2 The psychological analysis device 1 and the psychological analysis device 1 receive the inquiry information of the user through the line, receive the reply information of the operator 17a through the operator terminal 3 through the operator terminal 3, and digitize and analyze the psychological state of the user and the operators 17a and 17b. And an administrator terminal 19 that receives the psychological state of the operators 17a and 17b and the psychological state of the user who operates the user terminal 2.

  The psychological analysis device 1 provides emotion analysis data to the administrator 18 as a graph 7a based on a conversation exchanged between the user and the operator. The administrator screen of the administrator terminal 19 can be used as the output means of the graph 7a.

  The administrator 18 always grasps the psychological state of the user and the operators 17a and 17b displayed on the screen of the administrator terminal 19, and gives a relaxation instruction to the operator 17b who is troubled by the user's complaint processing. It can be transmitted from the terminal 19 to the emotion data providing apparatus 1.

  For example, when the tension of the operator 17b who is handling the user's complaint increases and only emotional response is possible, the operator terminal 3 is immediately switched, and the operator 17a operating the other operator terminal (not shown) performs the complaint process. A system for altering can be provided.

  The state 11 in which the user's mood of the user terminal 2 is corrected and the user is pleased is illustrated by an appropriate response of the operator 17a.

  In addition, the system which preserve | saves the emotion analysis data of operators 17a and 17b in a database, and can utilize for the counseling and management for operators is also object.

(Fifth embodiment)
As shown in FIG. 14, the emotion data providing apparatus according to the fifth embodiment of the present invention includes a plurality of operator terminals 3a, 3b, 3c for inputting operator reply information to inquiry information from the user through a line, User inquiry information is received from the user terminal 2, operator response information is received from the operator terminals 3a, 3b, and 3c through lines, and the psychological states of the user and a plurality of operators are digitized and analyzed, and the operator terminals 3a, 3b 3c is automatically selected and connected to the user terminal 2, and a psychological analysis device 1 is provided that allows the operator whose psychological state is stable to respond to the user.

  The psychological analysis device 1 executes the first stage correspondence with the user by the user terminal 2 and the answering horn system. At this time, the analysis results 7c, 7d, and 8a of the user's psychological state are graphed and displayed on the operator terminals 3a, 3b, and 3c.

  The psychological analysis apparatus 1 also analyzes the psychological states of the operators A, B, and C from the operator terminals 3a, 3b, and 3c, respectively. Based on each analysis result, relaxing means such as music and video can be transmitted to the user terminal 2 during the waiting time until the operator A who has a stable psychological state responds to the user, so as to calm the angry user. There is an advantage that you can.

  Furthermore, the operator A has an advantage that an appropriate response can be taken with the user by the line connection subjected to the matching process, and the waiting operator C can relax music, video, etc. during the waiting time in order to stabilize the psychological state. There is also an advantage that the means C can be received and care for the operator C can be performed.

(Sixth embodiment)
As shown in FIG. 15, the mobile phone according to the sixth embodiment of the present invention digitizes and analyzes the user's psychological state based on the five senses information including vision, hearing, smell, touch, and taste, When a user's psychologically depressed numerical value is detected based on the user history information stored in the information management module, an encouraging message is output with characters, pictures or sounds, and the user is psychologically uplifted Equipped with a psychological analysis device that outputs dream recall messages.

  The operation of the mobile phone will be described with reference to FIG. The psychological analysis device 1 (see FIG. 1) is system-on-chip and mounted on a mobile phone, and cannot be observed from the outside.

  The boss's mobile phone starts a conversation between the user and the boss in a calling step S20 for calling the user's mobile phone. The user's mobile phone receives the boss's “gummy” instruction, and the user's psychological state deteriorates to the “hi ...” state (step S21).

  The user's mobile phone transmits the user's psychological state to the address book corresponding to the contract telephone number to the server of the communication carrier company, and executes the look function step S22 for automatically setting the “Odachi” character. The psychological state may be transmitted while talking with the boss, or may be transmitted when the call with the boss is finished.

  The mobile phone of the user's acquaintance can see the user's psychological state before calling the user's mobile phone in the pre-calling step S23. That is, the acquaintance can check the psychological state of the other user through the address book before making a call. The address book may be provided in the server of the communication carrier company, or may be provided in the mobile phone of an acquaintance based on information automatically transmitted from the server of the communication carrier company by the address book update program.

  The user's psychological analysis and ecological color analysis can be matched to bring out the user's health status and to display on the display a message for maintaining health such as "You should eat carrots because your liver is weak" .

  When the acquaintance confirms the user's psychological state in step S22, the user knows the user's psychological state (step S24). Considering the phenomenon of “Ah, Odachiga!”, An encouraging message such as a text message “OK” or a healing image is transmitted to the user's mobile phone by e-mail (step S25).

  The encouragement message may be automatically displayed by the user's mobile phone as a text message such as “Do not fight”. In short, a program for analyzing the psychological state of the user and controlling emotions may be incorporated.

  The user has an advantage that the psychological state is stabilized by the mobile phone of the user who has received or displayed the encouragement message and can be shifted to a psychological state such as “Thank you”.

  With reference to FIG. 16, another sequence operation of the mobile phone will be described. The user's mobile phone incorporates a friend's program, and analyzes and analyzes the user's psychological state based on the five senses of information including vision, hearing, smell, touch, and taste. The five senses can be detected by a pressure sensor, a CCD camera, or an acceleration sensor.

  In step S26, the psychological state in which the user is depressed is shown. The ally program responds to the psychological state of the user (step S27), and displays an encouragement message that pushes the back to the user in the “very hard” state (step S28). For example, a text message such as “It ’s a great chance to change!” Should be displayed.

  If the user is psychologically upset, the mobile phone outputs a dream recall message (step S29). For example, the user may analyze the psychological state of “I want to be rich!” (Step S30) and display a text message such as “Today is rich!” As a dream recall message. The video corresponding to “It is rich!” Should be displayed on the display (step S31).

(Seventh embodiment)
The operator terminal 3d exemplified as the emotion data providing apparatus according to the seventh embodiment of the present invention displays a training screen as shown in FIG. In addition, description of the member or process which overlaps with 1st Embodiment mentioned above shall be abbreviate | omitted.

  The operator terminal 3d displays the psychological mark 9 on the screen in accordance with the psychological state of the user side and the operator side, displays the graph of the user data 7 and the graph of the operator data 8, and displays the simulation text 32 at the bottom of the screen. A message 33 from the manager is displayed at the bottom of the screen.

  The psychological graph 8 graphs, in real time, elements of anger, confusion, speech rate, appropriateness between speech, clarity of speech, ending, vocabulary, and intonation as operator data. The psychological graph 7 graphs, in real time, elements of anger, confusion, speech rate, appropriateness, speech clarity, ending, verbality, and intonation as user data.

  In the simulation text 32, as a user's question, “I want to purchase a new product A, how do I take a picture?” And as an operator's answer, “How to take a picture with the new product A? "Is displayed and the trainee is read out the operator part (OP).

The operator terminal 3d can display information that can be changed according to the manager's intention, such as knowledge, new product information, and a lot of complaints that are required by the company.
The message 33 displays, for example, text information such as “Relax! All right” so that the advice from the manager can be freely displayed according to the situation of the operator.

(Eighth embodiment)
The administrator terminal 19 exemplified as the emotion data providing apparatus according to the eighth embodiment of the present invention displays an administrator screen as shown in FIG. In addition, description of the member or process which overlaps with 4th Embodiment mentioned above shall be abbreviate | omitted.

  The administrator terminal 3d displays a danger signal 35, a desk number 36, a user state 37, an operator state 38, a work order icon 39a, and a seat order icon 39b on the screen. Further, the psychological state of the user and the operator can be displayed on the detailed screen 35a by clicking the eighth desk number 36 in the figure with a pointing device (for example, a mouse).

  Furthermore, by clicking the seat order icon 39b with a pointing device (for example, a mouse), the psychological state of the user and the operator can be converted into the screen display 39c in the operator's seat order. Similarly, by clicking the work order icon 39a with a pointing device (for example, a mouse), the psychological state of the user and the operator can be converted into a screen display in the order of the operator's work hours.

  The eighth disk number 36 in the figure is blinked to notify that the psychological state of the user or operator is dangerous. In FIG. 19, the second disk number 40 blinks a danger signal to notify that the psychological state of the user or operator is dangerous.

(Ninth embodiment)
As shown in FIG. 20, an operator terminal 3g exemplified as an emotion data providing apparatus according to the ninth embodiment of the present invention displays advice information from the emotion data providing apparatus on the screen. In addition, description of the member or process which overlaps with 4th Embodiment mentioned above shall be abbreviate | omitted.

  The operator terminal 3g generates a user-side emotion character 41, a user-side emotion barometer 42, a friend-side character 43, an operator-side emotion character 44, an operator-side emotion barometer 45, an answer / answer button 46, and a friend-side program on the screen. Comment 47 is displayed.

  The friend's program can provide multiple comments with different facial character expressions. For example, the display of the character 43a and the comment “The customer is angry! Let's calm down without drinking!” The display of the character 43b and the comment “Do your best! It's almost time to change!” The data corresponding to the psychological state of the operator is prepared for the display and the comment “Please calm down, please ...”.

(Tenth embodiment)
The user terminal 2 exemplified as the emotion data providing apparatus according to the tenth embodiment of the present invention responds to the voice access of the telephone user 11 as shown in FIG. Voice data transmitting means for automatically transmitting to the server 48, analysis data receiving means for receiving emotion analysis data obtained by digitizing and analyzing voice data from the server 48, and output means for outputting emotion analysis data to a display.

  The user terminal 2 can display the read emotion data in a graphic or text on the display.

  The server 48 analyzes the user's voice based on a large amount of emotion analysis data, compares it with the user's emotion history, performs matching processing, and uses the matched emotion data as analysis data for the user terminal 2 (for example, F-type). It has an emotion analysis engine 50 for automatic transmission to a portable terminal.

  Further, the server 48 can connect to the database server 49 and store the emotion history of the user. The user terminal 2, the server 48, and the database server 49 are connected via the Internet network to perform data transmission / reception.

  As shown in FIG. 22, the telephone server 48a automatically responds to a call from the mobile phone as the user terminal 2 of the telephone user 11 via the mobile phone switch, stores the voice data in the storage device, and the voice from the storage device. The data is read out and transferred to the application server 48b through the communication line. The application server 48b analyzes and compares the voice data based on the emotion analysis data stored in the emotion analysis engine 50, and the matching emotion analysis data is transmitted to the user through the mobile phone switch. A telephone user psychological analysis method for automatically transmitting to a mobile phone as the terminal 2 can be provided.

  The database server 49 can use means equivalent to the components shown in FIG.

  The telephone user psychological analysis method will be described with reference to FIG. 21, FIG. 22, and FIG. Microphone voice input step S53 for inputting the voice of the telephone user 11 to the mobile phone as the user terminal 2, storage step S54 for storing the voice input in the microphone voice input step S53 in the memory, and noise intensity mixed by the microphone voice input And determining whether or not the voice is valid as emotion determination data, and mail for transmitting the voice determined to be valid in the validity determination step S55 to the server 48 through the communication line as an attached file of an e-mail. The transmission step S56 and the server 48 extract parameters including the pitch, speed and timing of the voice, analyze and compare the voice based on the emotion analysis data stored in the emotion analysis engine 50 in advance (step S57), and the matching emotion Mobile phone as user terminal 2 through analysis line for analysis data Comprising the step S59 of transmitting, a display step S60 that the cellular phone as the user terminal 2 to display the graphic or text display emotion analysis data as the response data, the.

  The user terminal 2 processes the profile setting step S50 as a preset for telephone user psychological analysis. For example, the personality profile provided in the server 48 that becomes a selection factor such as response contents and wording is set, and the server 48 executes the response selection step S58.

  Further, the user terminal 2 selects any one of the trigger means for starting the application from the specific voice command, the generation of emotional relief or the key operation by the operation of the telephone user 11, and the mobile terminal main body as the user terminal 2 Save to memory.

  Furthermore, the user terminal 2 processes the response activation trigger setting step S52 by the operation of the telephone user 11. As a response, it is possible to select whether to display text information or bitmap image information on the display.

  The output history database (for example, the database server 49 in FIG. 21) refers to the personality profile, emotion (or state) determination result, and past data of the telephone user 11 in the output history database processing step S61, and determines the response type and expression method. Accumulate data to be determined.

(Eleventh embodiment)
In the psychological analysis method exemplified as the emotion data providing apparatus according to the eleventh embodiment of the present invention, as shown in FIG. 24, the administrator terminal 19 (see FIG. 18) selects the menu from the menu screen to the administrator. Step S62, a display step S63 for displaying a detailed screen, a step 64 for performing a simulation, a step S65 for simultaneously monitoring the psychological states of a plurality of users and operators, and a simulation prescription based on the simulation result Output step S66.

  The administrator terminal 19 can start the menu with a simple startup for the administrator. For example, the psychological analysis is performed by selecting any one of the menus of recruitment, training, management, implementation, data, and termination.

  In addition, the simulation prescription is: “Thank you for your hard work. If you speak a little more clearly and with a gentle tone, you will have a good impression.” Text analysis results graph of psychological state, user and operator feelings in GUI format The correlation table displayed in (5) should be output on the screen or to a printer.

(Twelfth embodiment)
As shown in FIG. 25, the mobile phone exemplified as the emotion data providing apparatus according to the twelfth embodiment of the present invention accesses the download site of the application software (for example, the viewing app), as shown in FIG. And step S68 for downloading application software (for example, the viewing app) and step S69 for making user settings for the application.

  In step S67, a download screen 70a is displayed on the display of the mobile phone to urge the user to start downloading Hyper Download. In step S68, a download completion screen 70b for notifying the download of Hyper View is displayed on the mobile phone display. In step S69, the pre-registration screen 70c is displayed.

  The pre-registration screen 70c displays a nickname input box, an i-αppli input box for inputting i-αppli automatic, manual or selection, and an input box for setting a response activation trigger, and the telephone user inputs arbitrary information to each input box. Input by button operation.

  With reference to FIG. 26, the operation of the mobile phone exemplified as the emotion data providing apparatus according to the twelfth embodiment will be described. The mobile phone launches the application program (viewer) after downloading the application program and completing the pre-registration process (step S71).

  Next, the cellular phone (for example, with a movie camera) executes a recording step S72 for recording the voice of the telephone user, and the cellular phone attaches the recording data to the e-mail and automatically transmits it to the server 48 (see FIG. 21). .

  Subsequently, the server 48 processes step S74 for analyzing the recording data, downloads the analysis result to the mobile phone, and performs step S75 to display the analysis result on the screen.

  In step S71, the hyperviewer's service menu 76a is displayed on the mobile phone display, and in step S72, the hyperviewer's instruction screen 76b is displayed on the mobile phone display to perform countdown processing. Within 30 seconds, the instruction word is inputted by voice and an e-mail is transmitted in step S73.

  In step S74, the screen 76c during emotion analysis is displayed on the display, and in step S75, the analysis result screen 76d is displayed on the display. An example analysis result may display the message "It seems that the previous life was a pig. Don't worry too much! Because it's an old story ...".

(Thirteenth embodiment)
The operator terminal exemplified as the emotion data providing apparatus according to the thirteenth embodiment of the present invention displays the customer center operation screen on the display of the operator terminal 3h as shown in FIG.

  The operator terminal 3h prompts the operator to input information that can change the size of the screen in accordance with the operation screen in the input box 77a, input box 77b, input box 77c, and input box 77d.

  In the detailed display 77 in the figure, a character display representing the psychological state of the user and the operator, a psychological state graph, and a message display are output.

  The operation of the operator terminal 3h (see FIG. 27) will be described with reference to FIG. The operator operates the operator terminal 3h and performs a name identification process step S78. For example, “Thank you for the call.

  The operator terminal 3h analyzes the voice input data and displays a psychological mark 9a and a graph 7a as a psychological analysis result on the display. If the operator is slowly and clearly known, a psychological analysis result exceeding the passing line can be obtained.

  Next, the operator terminal 3h executes a business knowledge handling process step S79. For example, in response to a question from the user, “I don't know how to use the QR code !?”, the operator asked the operator terminal 3h “How to use QR code? "Do you use?"

  When it is determined that the operator's voice input is stable from the beginning to the end, the operator terminal 3h displays the psychological mark 9b and the graph 7b on the display as psychological analysis results exceeding the pass line. Further, the operator terminal 3h analyzes voice data from the user and analyzes the user's psychological state. For example, if the user responds with a thank-you answer “Yes, I understand.”, Psychoanalyze that the user is responding with a good impression of the other person's position, and in step S80, the telephone call as the operator's voice input. Thank you very much. △△ was in charge of the appreciation "Thank you very much" psychological analysis, and displayed a psychological mark 9c with a relieved expression and a graph 7c of the psychological analysis result on the display.

  The operation of the operator terminal 3h (see FIG. 27) will be described with reference to FIG. The operator 82 operates the operator terminal 3h to execute the sound generation method master program. The operator terminal 3h guides the operator 82 (or the trainee) so that speech suitable for the operation can be made. For example, a psychological analysis result graph, a psychological mark, and Japanese and English character information are output as guidance information 81 on the display.

  The operator terminal 3h determines whether or not the operator 82 clearly pronounces the strength. For example, “Ae-Keko-Kakako” is inputted as a voice, the psychological state is digitized, and the psychological mark 9b and the psychological analysis result graph 7b are output to the display as analysis results.

  The operator terminal 3h determines whether or not the operator 82 controls the emotion and gives a description that the user is satisfied with. For example, when the operator terminal 3h receives a complaint phone 83 from the user saying “Hey! This product is!”, The operator terminal 3h changes the psychological mark 9b on the user side to an angry character and reports the event that the complaint phone has occurred. Notify operators and managers.

  In the operator terminal 3h, in response to the complaint call from the user in step S84, the operator's response is "I'm sorry. I'm sorry, can I ask you about the situation in more detail?" If it is, the psychological mark 9a and the psychological analysis result graph 7a are displayed on the display, and it is determined that the response is appropriate.

  As described above, according to the first to thirteenth embodiments of the present invention, for example, a user's psychological state analysis in a call center in which a plurality of operators are arranged, or two or more parties by matching processing of two or more parties. Providing psychological state analysis for analyzing the psychological state of each telephone user by letting other mobile phones talk at the same time, and providing a problem deciding means to respond on behalf of another operator when a problem occurs between the user and the operator it can.

  It should be noted that the actions and effects described in the embodiments of the present invention only list the most preferable actions and effects resulting from the present invention, and the actions and effects according to the present invention are described in the embodiments of the present invention. It is not limited to the ones.

  For example, the emotion data providing apparatus that inputs information on the five senses including the user's vision, hearing, smell, touch, and taste, analyzes the user's body, mind, and emotions, and outputs the analysis result with the information of the five senses. Can be applied.

  Emotional data provision device can be made into a neurochip and embedded in daily goods such as tableware to improve user satisfaction with information that appeals to the five senses such as “word”, “sight”, “taste”, etc. The satisfaction of the player embedded in the device can also be improved.

  The present invention can be applied to applications provided to telephone operator terminals, operator manager terminals, mobile phones, PDAs, car navigators, and the like.

  Also, it can be incorporated into a mobile phone to extract the user's emotions, psychology, preferences, and emotions from the mail content history, etc., and can be applied to applications that function as a consultant with a secretary function. It can also be applied to use of personal preferences accumulated in cards for related products.

The typical block diagram of the emotion data provision apparatus which concerns on the 1st Embodiment of this invention. The graph of the voiceprint data of the emotion data provision apparatus which concerns on the 1st Embodiment of this invention. The graph of the voiceprint data of the emotion data provision apparatus which concerns on the 1st Embodiment of this invention. The graph of the voiceprint data of the emotion data provision apparatus which concerns on the 1st Embodiment of this invention. The typical systematic diagram of the emotion data provision apparatus which concerns on the 1st Embodiment of this invention. The typical systematic diagram of the emotion data provision apparatus which concerns on the 1st Embodiment of this invention. The flowchart explaining operation | movement of the emotion data provision apparatus which concerns on the 1st Embodiment of this invention. The flowchart explaining operation | movement of the emotion data provision apparatus which concerns on the 1st Embodiment of this invention. The figure explaining the process outline | summary of the emotion data provision apparatus which concerns on the 1st Embodiment of this invention. The figure explaining the process outline | summary of the emotion data provision apparatus which concerns on the 1st Embodiment of this invention. The flowchart explaining operation | movement of the emotion data provision apparatus which concerns on the 2nd Embodiment of this invention. The typical systematic diagram of the emotion data provision apparatus which concerns on the 3rd Embodiment of this invention. The typical systematic diagram of the emotion data provision apparatus which concerns on the 4th Embodiment of this invention. The typical systematic diagram of the emotion data provision apparatus which concerns on the 5th Embodiment of this invention. The flowchart explaining operation | movement of the mobile telephone based on the 6th Embodiment of this invention. The flowchart explaining operation | movement of the mobile telephone based on the 6th Embodiment of this invention. The front view of the emotion data provision apparatus which concerns on the 7th Embodiment of this invention. The front view of the emotion data provision apparatus which concerns on the 8th Embodiment of this invention. The front view of the emotion data provision apparatus which concerns on the 8th Embodiment of this invention. The front view of the emotion data provision apparatus which concerns on the 9th Embodiment of this invention. Schematic system diagram of a mobile phone according to a tenth embodiment of the present invention. Schematic system diagram of a mobile phone according to a tenth embodiment of the present invention. The flowchart explaining operation | movement of the mobile telephone based on the 10th Embodiment of this invention. The flowchart explaining operation | movement of the emotion data provision apparatus which concerns on the 11th Embodiment of this invention. The flowchart explaining operation | movement of the emotion data provision apparatus which concerns on the 12th Embodiment of this invention. The flowchart explaining operation | movement of the emotion data provision apparatus which concerns on the 12th Embodiment of this invention. The front view of the emotion data provision apparatus which concerns on the 13th Embodiment of this invention. The flowchart explaining operation | movement of the emotion data provision apparatus which concerns on the 13th Embodiment of this invention. The figure explaining operation | movement of the emotion data provision apparatus which concerns on the 13th Embodiment of this invention.

Explanation of symbols

1 ... Psychological analysis device, 2 ... User terminal, 3 ... Operator terminal, 4a ... First line, 4b ... Second line, 48 ... Server, 49 ... Database server, 50 ... Emotion analysis engine.

Claims (15)

An operator terminal for inputting operator response information to the user inquiry information;
A psychological analysis device that receives user inquiry information from a user terminal through a first line, receives operator reply information from the operator terminal through a second line, and quantifies and analyzes the psychological state of the user and the operator; ,
Emotion data providing device comprising: The emotion data providing apparatus according to claim 1, further comprising psychological mark output means for converting the result of the analysis into a graphic user interface format and outputting the result. The emotion data providing apparatus according to claim 1, further comprising a graph output unit that converts the analysis result into a graphic format and outputs the result. The emotion data providing apparatus according to claim 1, wherein the psychological analysis apparatus outputs a result of the analysis to a terminal on an administrator side. The psychological analysis device converts the user's psychological state into a graphic format and outputs it to a plurality of the operator terminals, and outputs the user terminal to the operator terminals selected corresponding to the psychological states of the plurality of operators. 2. The emotion data providing apparatus according to claim 1, wherein automatic connection is made to match the inquiry information with the answer information. Means for digitizing and analyzing a psychological state of a telephone user calling through a telephone line; and psychological analysis transmitting means for transmitting the result of the analysis to a terminal of a group user whose telephone number is registered in an address book , A psychological analysis device. Emotions including psychological emotions, including psychological emotions, based on user history information stored in the user information management module, by analyzing the user's psychological state based on the five senses including visual, auditory, olfactory, tactile, and taste information Detects numerical values, outputs matching solution information, outputs an encouraging message with letters, pictures, or sounds when a depressed number is detected, and letters or pictures when the user is psychologically uplifted Alternatively, a mobile phone comprising a psychological analysis device that outputs the dream recall message using the five senses information or sound. A mobile phone that outputs emotion analysis data obtained by digitizing and analyzing first to third resonance frequency bands included in voice data of a telephone user to a display. 9. The mobile phone according to claim 8, wherein the display displays the emotion analysis data as a graphic or text. 9. The mobile phone according to claim 8, wherein the display analyzes and compares the voice data based on emotion analysis data accumulated in advance in an emotion analysis engine, and displays the matching emotion analysis data. A telephone server automatically responds to a call from a telephone user's mobile phone via a mobile phone switch, stores voice data in a storage device, reads the voice data from the storage device, and transfers the voice data to an application server through a communication line. The telephone user psychological analysis method for analyzing and comparing the voice data based on emotion analysis data accumulated in the emotion analysis engine by the application server and transmitting the matching emotion analysis data to the mobile phone through the mobile phone switch . A microphone voice input step for inputting the voice of the telephone user to the mobile phone;
A storing step of storing the voice input in the microphone voice input step in a memory;
An effective determination step of monitoring noise intensity mixed in by the microphone sound input and determining whether or not the sound is valid as emotion determination data;
A mail transmission step of transmitting the voice determined to be valid in the determination step to a server as an attachment file of an electronic mail through a communication line;
The server extracts parameters including pitch, speed, and timing of the voice, analyzes and compares the voice based on emotion analysis data stored in advance in the emotion analysis engine, and sends the matching emotion analysis data through the communication line. Transmitting to the mobile phone;
A display step in which the mobile phone displays the emotion analysis data on a display as a graphic or text; and
A display step in which the mobile phone displays emotion analysis data on a display as a graphic or text, and emotion analysis data that has been stored in advance in the emotion analysis engine by extracting parameters including voice pitch, speed, and timing in the mobile phone Incorporating a means for analyzing and comparing five sense information including voice and five senses (viewing angle, hearing, smell, and taste) and displaying matching emotion analysis data on the display; and
A telephone user psychological analysis method comprising: The telephone user psychological analysis method according to claim 12, wherein the emotion analysis data graphically displays emotions including joy, anger, sadness, discouragement, and excitement of the telephone user. The telephone user psychological analysis method according to claim 12, wherein the emotion analysis data is accumulated in an output history database provided in the server, and updated emotion analysis data is generated based on the past emotion analysis data. . 13. The method according to claim 12, further comprising: downloading from the server an application program for executing the microphone voice input step, the storage step, the validity determination step, the mail transmission step, and the display step on the mobile phone. Phone user psychology analysis method.

Images