JP2006011316A - Virtual conversational system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To perform a response in accordance with the characteristic of the conversation uttered by a user and the change of an external environment, etc. <P>SOLUTION: In the virtual conversational system, response data is determined by computing parameter values on the basis of input values of the characteristic of the voice of a speaker and the external environment, etc. Specifically, the system converts the characteristic of the user and the external environment, etc., into parameters and determines response data to meet the situation corresponding to these parameters. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a virtual conversation system that can virtually enjoy a conversation.

With the recent development of software technology, various virtual conversation systems have been developed. As an example of such a virtual conversation system, a voice conversation technology is used to decipher the user's voice, develop a scenario according to the decryption result, and enjoy a conversation with a virtual person. (See Non-Patent Document 1).
EZ Virtual Talk (Internet URL http://www.au.kddi.com/ezweb/contents/communication/index.html#EZ_VIRTUALTALK)

  However, in the conventional virtual conversation system, it is possible to select the response data in the scenario by using the recognition result of the conversation originated by the user. However, depending on changes in the user characteristics during the conversation, the external environment, etc. It was not possible to select response data for conversations made by users on an ad hoc basis.

  The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a virtual conversation system capable of performing a response in accordance with changes in user characteristics, external environment, and the like during conversation.

  In order to achieve the above object, according to the first invention of the present invention, the first parameter value is based on the data input from the user for determining the characteristics of the speaker in the virtual conversation system as the user's partner. Means for calculating the second parameter value based on data obtained via the network and affecting the characteristics of the speaker, means for outputting response data, and receiving the user's voice Means for calculating a third parameter value that affects speaker characteristics based on the received user voice, the first parameter value, the second parameter value, and the third parameter value. A first value is calculated by performing a predetermined calculation process on the stored first parameter value, second parameter value, and third parameter value, and means for storing the parameter value Means for recognizing the received user's voice, first response data, a plurality of voice data, a plurality of second values, and a plurality of second response data. A response database, the output response data corresponds to the first response data, the recognized user's voice corresponds to the voice data, and the calculated first value is the second value Means for retrieving second response data corresponding to the value from the response database, determining the retrieved second response data, and means for outputting the determined second response data. A virtual conversation system characterized by

  Further, according to the second invention of the present invention, means for calculating the first parameter value based on the data input from the user determining the characteristics of the speaker in the virtual conversation system that is the user's speaking partner, Means for calculating a second parameter value, means for outputting response data, means for receiving the user's voice, based on data obtained via the network and affecting the characteristics of the speaker; Based on the means for recognizing the face, the means for calculating a third parameter value that affects the characteristics of the speaker based on the received user's voice, and the feature of the face of the recognized user, Means for calculating a fourth parameter value affecting the characteristics of the speaker, and means for storing the first parameter value, the second parameter value, the third parameter value, and the fourth parameter value Means for calculating a first value by performing predetermined arithmetic processing on the stored first parameter value, second parameter value, third parameter value, and fourth parameter value; Means for recognizing the received user's voice; first response data; a plurality of voice data; a plurality of second values; and a response database associated with a plurality of second response data; The output response data corresponds to the first response data, the recognized user's voice corresponds to the voice data, and the calculated first value corresponds to the second value. And a means for determining the retrieved second response data and a means for outputting the determined second response data. Conversation system , It is.

  ADVANTAGE OF THE INVENTION According to this invention, a response can be performed according to the change of a user's characteristic in a conversation, an external environment, etc.

  Hereinafter, a virtual conversation system according to an embodiment of the present invention will be described with reference to the drawings.

  FIG. 1 is a diagram illustrating a computer that realizes a virtual conversation system according to an embodiment of the present invention.

  As shown in the figure, the virtual conversation system 1 includes a communication unit 11, a CPU 12, a memory 13, a voice output unit 14, a voice input unit 15, an input unit 16, an image input unit 17, and an HDD ( Hard disk drive) 18.

  The communication unit 11 is connected to the Internet and manages communication with the Internet.

  The CPU 12 controls the entire virtual conversation system, and implements a virtual conversation by executing a conversation processing program 21 described later.

  The memory 13 is used as a work area for executing the conversation processing program 21 or the like.

  The audio output unit 14 outputs audio from the virtual conversation system, and is, for example, a speaker.

  The voice input unit 15 is for inputting voice from the user, and is, for example, a microphone.

  The input unit 16 inputs parameter values for setting a virtual conversation partner in the virtual conversation system, and is an input device such as a keyboard.

  The image input unit 17 is for inputting a user image, and is, for example, an image input device such as an imaging camera.

  The HDD 18 stores a conversation processing program 21, a parameter value storage unit 22, a voice recognition program 23, an image recognition program 24, and a response data DB 25.

  The conversation processing program 21 realizes the operation of the virtual conversation processing system according to the embodiment of the present invention.

  The parameter value storage unit 22 is a basic parameter value input from a user, a network parameter value calculated from network data acquired from the outside via the Internet, a voice parameter value calculated from user voice data, and a user parameter An image parameter value calculated from image data is stored.

  The voice recognition program 23 performs voice recognition processing of voice input from the user via the voice input unit 15.

  The image recognition program 24 performs a process of recognizing the user image input via the image input unit 17.

  The response data DB 25 stores a plurality of scenarios. As shown in FIG. 14, this scenario has a plurality of response data, and this response data is associated with the previous response data, parameter values, and recognition data indicating the recognized speech.

  According to such a response data DB, the response data is specified by the previous response data, voice recognition data, and the overall parameter value.

  FIG. 2 is a block diagram showing functions of the virtual conversation system according to the embodiment of the present invention.

  As shown in the figure, the functional block of the virtual conversation system includes a basic parameter value receiving unit 31, a network parameter value calculating unit 32, a voice recognizing unit 33, a voice parameter value calculating unit 34, an image recognizing unit 35, and an image parameter value calculating. A unit 36, a response data determination unit 37, and a response data output unit 38.

  The basic parameter value receiving unit 31 receives basic settings input from the user, calculates basic parameter values, and outputs them to the response data determining unit 37. Here, the input basic settings are, for example, the speaker's age, sex, blood type, work, etc. in the virtual conversation system.

  The network parameter value calculation unit 32 calculates a network parameter value based on network data acquired from the Internet. The calculation of the network parameter value will be described later.

  The voice recognition unit 33 recognizes voice data input from the user.

  The voice parameter value calculation unit 34 calculates a voice parameter value based on the voice data recognized by the voice recognition unit 33. The calculation of the voice parameter value will be described later.

  The image recognition unit 35 recognizes the image of the user's face.

  The image parameter value calculation unit 36 calculates the image parameter value of the user's face image recognized by the image recognition unit 35. The calculation of the image parameter value will be described later.

  The response data determination unit 37 searches the response data DB 25 to determine response data based on the basic parameter value, the network parameter value, the voice parameter value, the image parameter value, and the recognized voice data. A method for determining the response data will be described later.

  The response data output unit 38 outputs the response data determined by the response data determination unit 37.

  FIG. 3 is a block diagram illustrating functions of the response data determination unit.

  As shown in the figure, the response data determination unit 37 includes an overall parameter value calculation unit 41 and a response data DB search unit 42.

  The total parameter value calculation unit 41 calculates a total parameter value by performing predetermined calculations on the input basic parameter value, network parameter value, audio parameter value, and image parameter value.

  The total parameter value is calculated by an equation as shown in FIG.

  Here, A1 to A4 are weighting coefficients. This weighting coefficient is set according to the type of scenario, and thus the speaker characteristics according to the type of scenario can be set.

  Next, each parameter value will be described.

  The basic parameter values are parameter values representing the speaker's age, sex, blood type, and work in the virtual conversation system. For example, the basic parameter value is set as shown in FIG.

  Here, the weighting coefficients B1 to B4 are set according to the type of scenario, and thus the speaker characteristics according to the type of scenario can be set.

  [Age] is set as follows.

0-10 years old [age] = 1
11-30 years old [age] = 2
31 to 70 years [age] = 3
71 years old ~ [age] = 4
For such parameter values, a table in which the input age and the [age] parameter value are associated is prepared (not shown), and the input age is used to search the table. Thus, the parameter value of [age] is calculated.

[Gender] is determined as follows: Male [Gender] = 1
Woman [Gender] = 1.5
For such parameter values, a table in which the entered gender and [gender] parameter values are associated is prepared (not shown), and the entered gender is used to search the table. Thus, the parameter value of [sex] is determined.

Also for [blood type] and [work], parameter values are determined according to each type. The setting items (age, sex, etc.) and parameter values for the basic parameter values can be arbitrarily set. For example, in addition to the above setting items, items relating to hobbies, items relating to personality, items relating to preferences, and the like may be provided.

  The network parameter value is a parameter value obtained by calculation from network data acquired on the network, and is a parameter as an external factor that affects the personality of the speaker.

  This network data is automatically acquired by the virtual conversation system, and may be acquired from a predetermined web site, may be acquired using a search engine, or may be set by a user. You may acquire from the front and the acquisition method does not ask | require.

  FIG. 7 is a diagram illustrating an expression for calculating the network parameter value. Here, C1 to C3 are weighting coefficients. This weighting coefficient is set according to the type of scenario, and thus the speaker characteristics according to the type of scenario can be set.

  The network data for the items [weather], [temperature], and [humidity] shown in FIG. 7 are acquired from, for example, a weather forecast site. The network parameter value [weather] is set as follows.

Sunny [Weather] = 1
Cloudy [Weather] = 2
Rain [weather] = 3
For such parameter values, a table in which the data indicating the weather and the parameter values of [weather] are associated is prepared (not shown), and the table is obtained using the acquired data indicating the weather. By searching, the parameter value of [weather] is determined.

  [Temperature] is set as follows.

~ 5 ℃ [Air temperature] = 1
6 ℃ ~ 20 ℃ [Air temperature] = 2
21 ℃ -30 ℃ [Temperature] = 3
31 ° C ~ [Air temperature] = 4
For such parameter values, a table in which the data indicating the air temperature and the parameter value of [air temperature] are associated (not shown) is prepared, and the table is obtained using the acquired data indicating the air temperature. By searching, the parameter value of [temperature] is determined.

  As for [humidity], a parameter value is determined according to the type of humidity.

  The setting items (weather, temperature, etc.) and parameter values of the network parameter value can be arbitrarily set. For example, in addition to the setting items, items related to hobbies (for example, baseball game results) may be provided.

  The voice parameter value is a parameter value representing the voice of the speaker in the virtual conversation system. The voice parameter value is determined based on the voice strength, speed, and ending. In other words, the input speech is recognized by the speech recognition process for the strength, speed, and ending of the speech, and is strong, weak, or quite strong against the reference speech strength, speed, and ending (strong) ), Whether it is fairly weak (weak), or similar to the standard (standard), and using the table as shown in FIG. 8, the parameter values for the voice strength, speed, and ending are respectively It is determined.

  The input voice parameter value is calculated by the formula shown in FIG. Here, D1 to D3 are weighting coefficients. This weighting coefficient is set according to the type of scenario, and thus the speaker characteristics according to the type of scenario can be set.

  FIG. 10 is a diagram showing a table for determining the final audio parameter value. In the figure, the predicted parameter value is the previous final speech parameter value, and the input speech parameter value is the parameter value determined by the equation shown in FIG. The final audio parameter value is calculated according to the following rules.

  1. When the difference between the predicted parameter value and the input voice parameter value is 2 or less, the input voice parameter value becomes the final voice parameter value.

  2. When the difference between the prediction parameter value and the input speech parameter value is +2 or more, a value obtained by adding “2” to the prediction parameter value is the final speech parameter value.

  3. When the difference between the predicted parameter value and the input speech parameter value is −2 or less, a value obtained by adding “−2” to the predicted parameter value is the final speech parameter value.

  By determining the voice parameter value by such a method, it is possible to prevent the final voice parameter value from abruptly fluctuating and to consider time series elements.

  The image parameter value is a parameter value indicating a user state that affects a speaker in the virtual conversation system. Here, the user's state is converted into a parameter value by paying attention to the face image, the eye image, the mouth image, the whole body image, etc., among the image images of the user recognized. A method for determining the state of the user is arbitrarily determined.

  In this embodiment, the image recognition process determines whether the image corresponds to “Wake up”, “Wake up”, “Sleeping”, or “Sleeping”. Of the recognized user images, the determination is made by paying attention to the face image, the eye image, the mouth image, the whole body image, and the like.

  The input image parameter value corresponding to the determined user state is determined with reference to the table shown in FIG. Then, the final image parameter value is obtained according to the following rules.

  In the present embodiment, it is assumed that the image recognition parameter value is acquired every 10 seconds, but the present invention is not limited to this.

  1. When the input image parameter value is “0” or “1”, the sum of the parameter values for one minute including the current input image parameter value is obtained, and the table shown in FIG. 12 is searched from the obtained parameter value. To determine the final image parameter value.

  2. If the input image parameter value is “2” and 10 seconds ago is “0” or “1”, the final image parameter is set to “2”.

  3. If the input image parameter value is “2” and 10 seconds ago is “2” or “3”, the final image parameter is set to “3”.

  4). When the input image parameter value is “2”, the final image parameter is “3”.

  FIG. 13 is a diagram illustrating an example of the final image parameter value. In the figure, an image parameter value acquired at 60 seconds is an acquired image parameter value, and the acquired image parameter value is rewritten to a final image parameter value according to the above rule. ing. It is assumed that the calculation formulas and tables for calculating each parameter value described above are stored in the HDD 18 (not shown).

  In the present embodiment, an example has been described in which the parameter value indicating sleepiness is calculated based on a predetermined part (mouth, eyes, whole body) in the acquired user image, but the present invention is not limited to this. Instead, information based on the user's personality (for example, desired color) is acquired from the acquired user's image and replaced with the parameter value, and is included in the scope of the present invention.

  The response data DB search unit 42 determines the response data by searching the response data DB 25 based on the recognized voice data and the total parameter value calculated by the total parameter value calculation unit 41.

  FIG. 4 is a diagram illustrating an example of one scenario among a plurality of scenarios stored in the response data DB.

  As shown in the figure, the scenario I has a plurality of response data A-1 to A-10, and these response data A-1 to A-10 are associated with each other. In the figure, the arrows indicate the flow of time. For example, after outputting the response data A-1, the arrows indicate that the voice from the user is received, and then the response data A-2 is output. After the response data A-2 is output, the voice from the user is received, and the response data A-2 or the response data A-3 is output depending on the recognized voice and the result of the comprehensive parameter value. .

  FIG. 14 is a diagram illustrating data stored in the response data DB. As shown in the figure, a plurality of voice recognition data is associated with one response data. Each voice recognition data is associated with a plurality of parameter values, and response data is associated with the parameter values.

  That is, by using the response data, the speech recognition data (recognized speech data) and the parameter values that have been output previously, it is possible to search for response data corresponding to the speech input by the user.

  Next, the operation of the virtual conversation system according to the embodiment of the present invention will be described with reference to the flowchart of FIG.

  When the virtual conversation system is activated, first, a basic setting for determining basic characteristics of a speaker who is a user's partner input from the user is accepted (S1). Here, the basic parameter values are the speaker's age, sex, blood type, work, etc. in the virtual conversation system.

  Next, basic parameter values are calculated for the input basic settings (S2). As described above, the basic parameter value is calculated by searching a table (not shown) in which the input basic setting is associated with the basic parameter value.

  Next, it is determined whether it is time to acquire data (network data and image data) related to external factors other than the basic parameter values (S3). The acquisition time can be arbitrarily set by the user. For example, the network data can be arbitrarily set at 9 o'clock in the morning and the user image every 10 seconds.

  If it is determined in S3 that it is time to acquire data, it is next determined whether it is time to acquire network data (S4). If it is determined in S4 that it is time to acquire network data, network data is acquired. The network data acquisition method is not particularly limited as described above. In the present embodiment, it is assumed that network data for the items [weather], [temperature], and [humidity] are acquired from the weather forecast site set by the user (S5).

  Next, network parameter values are calculated for the acquired network data (S6). As described above, the network parameter value is calculated by searching a table (not shown) in which the network data relating to the items [weather], [temperature], and [humidity] are associated with the network parameter value. (S6).

  After the network parameter value is calculated in S6 or when it is determined in S4 that it is not the network data acquisition time, it is determined whether or not it is the image data acquisition time (S7).

  If it is determined in S7 that it is time to acquire image data, image data indicating the user's image is acquired (S8), and image recognition processing is performed on the acquired image (S9). Then, as described above, the image parameter value is calculated for the user image that has undergone this image recognition process (S10), and the process returns to S3.

  On the other hand, if it is determined in S3 that it is not the data acquisition time, the response data is output as a sound (S21). Next, the user's voice corresponding to the voice response data output in S21 is acquired (S22), the acquired voice is recognized (S23), and the voice parameter value is calculated (S24). ).

  Next, a total parameter value is calculated based on the calculated basic parameter value, network data parameter value, image parameter value, and audio parameter value (S25).

  Thereafter, the response data to be output next is determined by searching the scenario database based on the response data output last time, the recognized voice data, and the total parameter value calculated in S25 (S26).

  Thereafter, a determination is made as to whether the conversation has ended. The determination as to whether or not the conversation has ended is made when the speech recognition data (recognized voice data) corresponding to the output response data does not exist in the response data database. If it is determined in S27 that the conversation has ended, the process in the conversation system is ended. On the other hand, if it is determined that the conversation has not ended, the process returns to S3.

  Therefore, according to the virtual conversation system according to the embodiment of the present invention, the parameter value of the speaker in the virtual conversation system changes according to the external factor, and as a result, the scenario according to the external factor can be developed. .

  Note that the present invention is not limited to the above-described embodiment as it is, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage. In addition, various inventions can be formed by appropriately combining a plurality of components disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, constituent elements over different embodiments may be appropriately combined.

  In addition, the method described in the embodiment is, for example, a magnetic disk (floppy (registered trademark) disk, hard disk, etc.), optical disk (CD-ROM, DVD) as programs (software means) that can be executed by a computer (computer). , MO, etc.), a semiconductor memory (ROM, RAM, flash memory, etc.) or the like, or can be transmitted and distributed via a communication medium. The program stored on the medium side includes a setting program that configures software means (including not only the execution program but also a table and data structure) in the computer. A computer that implements this apparatus reads a program recorded on a recording medium, constructs software means by a setting program as the case may be, and executes the above-described processing by controlling the operation by this software means. The recording medium referred to in this specification is not limited to distribution, but includes a storage medium such as a magnetic disk or a semiconductor memory provided in a computer or a device connected via a network.

It is a figure which shows the computer which implement | achieves the virtual conversation system which concerns on embodiment of this invention. It is a block diagram which shows the function of the virtual conversation system which concerns on embodiment of this invention. It is a block diagram which shows the function of a response data determination part. It is a figure which shows the example of one scenario among the several scenarios stored in response data DB. It is a figure which shows the type | formula for calculating a comprehensive parameter value. It is a figure which shows the type | formula for calculating a basic parameter value. It is a figure which shows the type | formula for calculating a network parameter value. It is a table for calculating | requiring an audio | voice parameter value. It is a figure which shows the type | formula for calculating an input audio | voice parameter value. It is a figure which shows the table for determining the last audio | voice parameter value. It is a figure which shows the table for determining an input image parameter value. It is a figure which shows the table for determining a final image parameter value. It is a figure for demonstrating how to obtain | require a final image parameter value. It is a figure which shows the data stored in response data DB. It is a flowchart for demonstrating operation | movement of the virtual conversation system which concerns on embodiment of this invention.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Virtual conversation system, 10 ... Bus, 11 ... Communication part, 12 ... CPU, 13 ... Memory, 14 ... Voice output part, 15 ... Voice input part, 16 ... Input part, 17 ... Image input part, 18 ... HDD ( Hard disk drive), 21 ... conversation processing program, 22 ... parameter value storage unit, 23 ... voice recognition program, 24 ... image recognition program, 25 ... response data DB.

Claims (7)

Means for calculating a first parameter value based on data input from the user for determining the characteristics of the speaker in the virtual conversation system that is the user's speaking partner;
Means for calculating a second parameter value based on data obtained via the network and affecting the characteristics of the speaker;
Means for outputting response data;
Means for receiving the user's voice;
Means for calculating a third parameter value that affects speaker characteristics based on the received user voice;
Means for storing the first parameter value, the second parameter value, and the third parameter value;
Means for calculating a first value by performing a predetermined calculation process on the stored first parameter value, second parameter value and third parameter value;
Means for recognizing the received user voice;
A response database in which the first response data, the plurality of audio data, the plurality of second values, and the plurality of second response data are associated;
The output response data corresponds to the first response data, the recognized user's voice corresponds to the voice data, and the calculated first value corresponds to the second value. Means for retrieving second response data from the response database and determining the retrieved second response data;
A virtual conversation system, comprising: means for outputting the determined second response data. Means for calculating a first parameter value based on data input from the user for determining the characteristics of the speaker in the virtual conversation system that is the user's speaking partner;
Means for calculating a second parameter value based on data obtained via the network and affecting the characteristics of the speaker;
Means for outputting response data;
Means for receiving the user's voice;
Means for recognizing the user's face;
Means for calculating a third parameter value that affects speaker characteristics based on the received user voice;
Means for calculating a fourth parameter value that affects speaker characteristics based on the recognized facial features of the user;
Means for storing the first parameter value, the second parameter value, the third parameter value and the fourth parameter value;
Means for calculating a first value by performing predetermined arithmetic processing on the stored first parameter value, second parameter value, third parameter value, and fourth parameter value;
Means for recognizing the received user voice;
A response database in which the first response data, the plurality of audio data, the plurality of second values, and the plurality of second response data are associated;
The output response data corresponds to the first response data, the recognized user's voice corresponds to the voice data, and the calculated first value corresponds to the second value. Means for retrieving second response data from the response database and determining the retrieved second response data;
A virtual conversation system, comprising: means for outputting the determined second response data.   The first parameter value is a parameter value indicating a speaker's personal attribute, and the second parameter value is a parameter value based on an external factor different from the personal attribute. 2. The virtual conversation system according to 2.   3. The virtual conversation system according to claim 1, wherein the third parameter value is determined based on the received user's voice strength, speed, and ending voice level. 4.   The third parameter value is determined based on the received user's voice strength, speed, ending voice level, and a previously calculated third parameter value. The virtual conversation system according to claim 1 or claim 2. In a virtual conversation program executed in a virtual conversation system including a response database in which first response data, a plurality of audio data, a plurality of second values, and a plurality of second response data are associated with each other,
Means for causing a computer to calculate a first parameter value based on data input from a user for determining characteristics of a speaker in a virtual conversation system as a user's partner;
Means for causing the computer to calculate a second parameter value based on data obtained via the network and affecting the characteristics of the speaker;
Means for causing the computer to output response data;
Means for causing the computer to receive the user's voice;
Means for causing a computer to calculate a third parameter value that affects the characteristics of the speaker based on the received user voice;
Means for causing a computer to calculate a first value by performing predetermined arithmetic processing on the calculated first parameter value, second parameter value, and third parameter value;
Means for causing a computer to recognize the received user voice;
The output response data corresponds to the first response data, the recognized user's voice corresponds to the voice data, and the calculated first value corresponds to the second value. Means for retrieving second response data from the response database and causing the computer to determine the retrieved second response data;
A virtual conversation program comprising: means for causing the computer to output the determined second response data. In a virtual conversation program executed in a virtual conversation system including a response database in which first response data, a plurality of audio data, a plurality of second values, and a plurality of second response data are associated with each other,
Means for causing a computer to calculate a first parameter value based on data input from a user for determining characteristics of a speaker in a virtual conversation system as a user's partner;
Means for causing the computer to calculate a second parameter value based on data obtained via the network and affecting the characteristics of the speaker;
Means for causing the computer to output response data;
Means for causing the computer to receive the user's voice;
Means for causing the computer to recognize the user's face;
Means for causing a computer to calculate a third parameter value that affects the characteristics of the speaker based on the received user voice;
Means for causing the computer to calculate a fourth parameter value that affects the characteristics of the speaker based on the recognized facial features of the user;
Means for causing a computer to calculate a first value by performing a predetermined calculation process on the first parameter value, the second parameter value, the third parameter value, and the fourth parameter value;
Means for causing a computer to recognize the received user voice;
The output response data corresponds to the first response data, the recognized user's voice corresponds to the voice data, and the calculated first value corresponds to the second value. Means for retrieving second response data from the response database and causing the computer to determine the retrieved second response data;
A virtual conversation program comprising: means for causing the computer to output the determined second response data.

Images