JP4970330B2 - Conversation support apparatus and method - Google Patents

Description

  The present invention relates to a technique for supporting a conversation via a communication network.

  Examples of the conversation performed via the communication network include a video conference and a chat performed on the Internet.

  Even if the user wants to activate the conversation, the user may be troubled because good topics and opinions are not raised. Such a situation tends to occur when, for example, a clue for a speech is not found or a previous conversation with a partner is forgotten.

  For example, there are techniques disclosed in Patent Documents 1 and 2 as techniques that are conscious of such problems.

  According to Patent Document 1, a profile of a partner and a topic that can be shared with the partner are registered in advance, and pre-registered information (a topic that can be shared with the profile of the partner) is displayed to the user during a conversation. (See, for example, FIGS. 1 and 5).

  According to Patent Literature 2, when a user inputs a condition (gender, age, etc.) related to a partner, a partner profile that matches the condition is searched. Then, keywords are extracted from the retrieved partner profile, and content specified based on the extracted keywords is provided to the user (see, for example, FIGS. 1, 4 and 5).

JP 2001-282687 A JP 2003-223406 A

  However, in Patent Document 1, information displayed to the user is pre-registered information. In Patent Document 2, the content provided to the user is content based only on the searched partner profile.

  In short, in both Patent Documents 1 and 2, the topic to be presented is fixed for each partner. Therefore, the same topic is presented to the user regardless of changes in the situation, such as a long conversation with the other party or multiple conversations with the same party.

  Therefore, an object of the present invention is to change a speech trigger with the same partner to be presented to the user to a different speech trigger according to a change in the situation.

  A conversation support device that supports a conversation performed via a communication network between a first user using a first user terminal and a second user using a second user terminal is constructed.

  The conversation support apparatus according to the first aspect extracts a character string from the second user related data related to the second user, and converts the character string into another character string related thereto. And the speech trigger with respect to the 2nd user of a 1st user is created based on the converted character string, and the created speech trigger is transmitted to a 1st user terminal. As a result, a speech trigger based on the character string extracted from the second user-related data and converted, which is a speech trigger for the second user of the first user, is displayed on the display screen of the first user terminal.

  The conversation support device according to the second aspect extracts a character string including an added or updated portion of the second user related data related to the second user. And the speech trigger with respect to the 2nd user of a 1st user is created based on the extracted character string, and the created speech trigger is transmitted to a 1st user terminal. Accordingly, the display trigger of the first user terminal is a speech trigger for the second user of the first user, and the text trigger including the added or updated portion of the second user related data is based on the speech trigger. Is displayed.

  FIG. 1 shows a computer system according to an embodiment of the present invention.

  A plurality of user terminals 50, a conversation space providing server 100, and a general data source device 70 are connected to the communication network 60.

  The user terminal 50 is a computer (for example, a personal computer or a mobile phone) used by the user. The user terminal 50 includes, for example, a network interface for communicating via an input device (for example, a keyboard or pointing device) 51, a display device 53, a CPU (Central Processing Unit) 55, a storage device 57, and a communication network 60. And a device (network I / F) 59.

  A conversation space providing server (hereinafter abbreviated as “server”) 100 is a system composed of one or a plurality of computers, and includes, for example, a CPU 107, a storage device 105, and a network I / F 111. The storage device 105 is, for example, a main storage device (for example, a memory) and / or an auxiliary storage device (for example, a drive of a nonvolatile storage medium). The server 100 provides a conversation space for the user terminal 50. A “conversation space” is a virtual space in which a conversation takes place.

  The general data source device 70 is a device that is a source of general data (typically, a device that stores general data), and specifically, for example, at various sites connected to the communication network 60. is there. The “general data” referred to in the present embodiment is data other than the conversation log data and user profile data described later, and is added, changed, or changed regularly or irregularly in the same manner as the conversation log data and user profile data. The data to be deleted. Examples of general data include news data issued by a newspaper company, blog data updated by a user, and the like.

  FIG. 2 is an explanatory diagram of the outline of the present embodiment.

  Each object representing each user who is logged in to the server 100 exists in the conversation space 900. The object is, for example, an avatar. According to the conversation space 900 illustrated in FIG. 2, the first user avatar 801 and the second user avatar 803 exist. The view of the conversation space 900 of FIG. 2 is a view displayed on the terminal 50 of the first user based on the position of the first user avatar 801.

  When the first user avatar 801 moves (see dotted arrow 805) and the distance between the first user avatar 801 and the second user avatar 803 is reduced to a predetermined distance or less, the first user and the second user It becomes possible to talk. At that time, in the view of the conversation space 900 for the first user, a speech trigger window (hereinafter abbreviated as “window”) 950 is developed in the vicinity of the second user avatar 803. This window 950 displays a conversation trigger with the second user. The speech trigger is, for example, a topic keyword and / or a question sentence for the second user. The first user can speak to the second user using a desired speech trigger (for example, “Makuhari XX event” as shown).

  One feature of the present embodiment is that the speech trigger displayed in the window 950 is a speech trigger corresponding to a change in the situation. For example, the speech trigger displayed on the window 950 is changed according to the progress of the conversation between the first user and the second user. Specifically, for example, as shown in FIG. 3, when a keyword “reiko” or “travel” is included in a remark from the second user, a replay trigger related to “reiko” or “travel” (for example, “Inu Kashiwazaki” and “Which would you like to travel?” Are given high priority and displayed on the window 950 (that is, the display on the window 950 changes from (A) to (B)). As a result, the first user can develop a conversation about “lion” or “travel”.

  Hereinafter, with reference to FIG. 2, FIG. 4, FIG. 5 and FIG. 6, the server 100 configured to be able to present a speech trigger according to a change in the situation to the first user will be described. In the following description, the first user is taken as an example of the information providing destination from the server 100, but the description is also applicable to the second user who is the partner of the first user's conversation (for example, The first user is presented with a speech trigger with the second user who is the counterpart of the first user, but similarly, the second user is presented with a speech trigger with the first user).

  The server 100 includes a conversation space providing unit 501, a conversation support unit 503, a profile database (profile DB) 1011, and a conversation log database (conversation log DB) 1013. The conversation space providing unit 501 and the conversation support unit 503 are computer programs executed by the CPU 107, for example. The profile DB 1011 and the conversation log DB 1013 are stored in the storage device 105 (or a remote storage device outside the server 100). The profile DB 1011 includes profile data for each user. In the profile data, character strings corresponding to various attribute items related to the user's profile (hereinafter referred to as “keyword” for convenience) are described (for example, if the attribute item is “Birthplace”, the keyword is “Yamagata Prefecture”). ”). The conversation log DB 1013 includes conversation log data for each user. In the conversation log data, a conversation log for each conversation performed by the user corresponding to the conversation log data is recorded. Each conversation log includes, for example, the length of the conversation, the text for each utterance, and the user ID that made the utterance.

  The conversation space providing unit 501 transmits information related to the conversation space 900 of the view based on the position of the first user avatar 801 to the first user terminal 50. The information includes, for example, information about the conversation space 900 itself, information about the first user avatar 801, and information about the second user avatar 803 (for example, a position in the conversation space 900). A WEB browser (for example, a dedicated viewer is plugged in) 501 executed by the CPU 55 in the user terminal 50 is based on the information from the conversation space providing unit 501, and the first user avatar 801 and the second user avatar 803. Is displayed on the display device 53.

  In response to an operation from the first user, the conversation space providing unit 501 moves the first user avatar 801 in the conversation space 900 as indicated by a dotted arrow 805, for example. When the conversation space providing unit 501 detects that the distance between the first user avatar 801 and the second user avatar 803 is equal to or less than a predetermined distance, the conversation space providing unit 501 closes the second user avatar 803 to the second user of the first user. A window 950 displaying a speech trigger is expanded to accept speech input from the first user. Specifically, for example, the conversation space providing unit 501 notifies the conversation support unit 503 of the start of the conversation between the first user and the second user, and in response to the notification, a speech trigger to be presented to the first user. Received from the conversation support unit 503. When the conversation space providing unit 501 detects the end of the conversation between the first user and the second user (for example, one of the first user avatar 801 and the second user avatar 803 moves and a distance between them is predetermined). When it is detected that the user is away from the distance), the conversation support unit 503 is notified of the end of the conversation. In addition, the conversation space providing unit 501 has a conversation log corresponding to the conversation (corresponding to the first user) every time the first user or the second user speaks until the conversation is started and ended. The message is added to the conversation log in the conversation log data.

  In response to the conversation start notification from the conversation space providing unit 501, the conversation support unit 503 extracts the speech trigger to be presented to the first user from the latest profile DB 1011, the latest conversation log 1013, and the latest general data. It generates based on the keyword, and outputs the generated speech trigger to the conversation space providing unit 501. Therefore, the speech trigger output here is a speech trigger according to the latest status of the second user. That is, even if the second user is not the first partner, if the situation has changed from the previous conversation with the second user, a speech trigger different from the speech trigger presented at the previous conversation is It can be presented to the first user.

  The conversation support unit 503 has a function of detecting changes in the status of the profile DB 1011, conversation log DB 1013, and general data. Each time the conversation support unit 503 detects a change in the situation, the conversation support unit 503 generates a speech trigger according to the change in the situation and outputs it to the conversation space providing unit 501. As a result, during a conversation, a speech trigger corresponding to a change in the situation is displayed in the window 950 instead of the speech trigger that was displayed before the situation change.

  Hereinafter, the function of the conversation support unit 503 will be described in more detail with reference to FIG.

  FIG. 4 shows functional blocks of the conversation support unit 503.

  The conversation support unit 503 includes a familiarity table 1027, a familiarity monitoring unit 1028, a partner information extraction unit 1029, a change detection unit 1031, a common term extraction unit 1033, a speech extraction unit 1035, a keyword conversion unit 1037, a keyword extraction unit 1039, weighting An adjustment unit 1041, an output unit 701, and a selection detection unit 1043 are included.

  The familiarity table 1027 records the familiarity indicating the degree of familiarity between users for each user. The familiarity between users is determined based on, for example, the total conversation time length and / or the number of conversations between the users. More specifically, for example, if there are three levels of intimacy and the total conversation time length is less than 10 hours or the number of conversations is less than 3, the familiarity is set to the minimum value “1”, and the total conversation time length Is 30 or more and the number of conversations is 5 or more, the maximum value is "3", otherwise it is "2".

  The intimacy monitoring unit 1028 monitors the conversation log DB 1013 and updates the intimacy table 1207 based on the monitoring result (for example, the latest total conversation time length and / or the number of conversations between users).

  The partner information extraction unit 1029 extracts a keyword corresponding to a specific attribute item in the second user profile data (in other words, the partner profile data), and sends the name and keyword of the specific attribute item to the keyword conversion unit 1037. Here, the “specific attribute item” is an attribute item according to the familiarity between the second user and the first user. Specifically, for example, there are [sex], [name], [hometown], etc. as items according to the familiarity “1”, and [family composition] as items according to the familiarity “2”. There are [Religion] and [Supporting Political Party] as items according to intimacy “3”. The partner information extraction unit 1029 extracts a keyword corresponding to an item below the familiarity between the second user and the first user. For example, if the closeness is “2”, the partner information extraction unit 1029 selects keywords corresponding to [gender], [name], [birthplace], etc. in addition to keywords corresponding to [family composition]. Extract.

  For example, the change detection unit 1031 periodically checks the profile DB 1011 to detect that a keyword corresponding to an attribute item among a plurality of attribute items in the second user profile data has changed. The change detection unit 1031 extracts the changed keyword from the second user profile data, and sends the keyword and its attribute item name to the keyword conversion unit 1037. In addition, when the changed keyword is a keyword of an item higher than the intimacy between the first user and the second user, the changed keyword is not acquired.

  The common term extraction unit 1033 uses a keyword common to a plurality of attribute items of the first user profile data and the second user profile data (for example, “male” for “sex”, “Yamagata” for “hometown”. And the extracted keyword and its attribute item name are sent to the keyword conversion unit 1037.

  The speech extraction unit 1035 extracts words having a high frequency of occurrence (for example, words having a predetermined number or more per unit character) as keywords from the conversation log data corresponding to the second user. The speech extraction unit 1035 changes the destination of the extracted keyword depending on whether the conversation log from which the keyword is extracted is a conversation log with the first user or a conversation log with another user. Specifically, the keyword extracted from the log of the conversation with the first user is sent to the weighting adjustment unit 206. On the other hand, the keyword extracted from the log of the conversation with the user other than the first user is sent to the keyword conversion unit 1037 (however, if the conversation is also in which the first user is participating, the extracted keyword is Sent to the weight adjustment unit 206). The keyword sent to the weight adjustment unit 206 may be presented to the first user as it is, but the keyword sent to the keyword conversion unit 1037 is weighted as another keyword related to the keyword. It is sent to the adjustment unit 206 and then presented to the first user. Therefore, it is possible to prevent the first user from guessing the content of the conversation in which the first user has not participated.

  When the keyword conversion unit 1037 receives a keyword, the keyword conversion unit 1037 converts the received keyword into another keyword related to the keyword, and converts the converted keyword and the source of the keyword (received keyword) as a basis thereof. The type of the data that has been changed is sent to the weight adjustment unit 1041. Specifically, for example, if the received keyword is a keyword from the partner extraction unit 1029, the change detection unit 1031 or the common term extraction unit 1033, it is determined that the source data type is the second user profile data, If it is a keyword from the extraction unit 1035, it is determined that the source data type is conversation log data. As a method for converting the received keyword into another keyword related thereto, an existing method such as a method disclosed in Japanese Patent Laid-Open No. 5-282367 may be employed.

The keyword extraction unit 1039 extracts keywords from the general data, and sends the extracted keywords to the weight adjustment unit 1041. Here, the general data from which the keyword is extracted is the general data searched by the keyword extraction unit 1039. The keyword extraction unit 1039 includes the following six types of keywords:
(1) a keyword sent from the partner information extraction unit 1029 to the keyword conversion unit 1037;
(2) a keyword sent from the change detection unit 1031 to the keyword conversion unit 1037;
(3) a keyword sent from the common term extraction unit 1033 to the keyword conversion unit 1037;
(4) a keyword sent from the speech extraction unit 1035 to the keyword conversion unit 1037;
(5) a keyword sent from the speech extraction unit 1035 to the weighting adjustment unit 1041;
(6) a keyword sent from the keyword conversion unit 1037 to the weighting adjustment unit 1041;
The general data related to at least one of them is searched from the general data source device 70, and keywords are extracted from the searched general data. Since the general data may be changed at a timing that does not depend on the conversation between the first and second users, the keyword extraction unit 1039 periodically relates to at least one of the six types of keywords. Keywords may be extracted by searching general data.

  The weighting adjustment unit 1041 receives keywords from the keyword conversion unit 1037, the speech extraction unit 1035, and the keyword extraction unit 1039, and uses the received keyword itself as a topic keyword or a question sentence created based on the received keyword. Create a speech trigger. At that time, the weight adjustment unit 1041 determines the priority order of the speech triggers, and sends each speech trigger and its priority order to the output unit 701.

  The output unit 701 receives each speech trigger and the priority order of each speech trigger from the weight adjustment unit 1041, and transmits each speech trigger to the WEB browser 501 according to the priority order. As a result, the speech triggers arranged in the descending order of priority are displayed in the window 950 by the WEB browser 501. Each speech trigger displayed in the window 950 can be selected by the first user. The speech trigger selected by the first user is reflected on a speech input screen (input screen for conversation) (not shown) of the first user, for example. Thereby, the labor of inputting the speech trigger can be saved.

  The selection detection unit 1043 detects a speech trigger selected on the window 950 (for example, a speech trigger specified using a pointing device), and notifies the weighting adjustment unit 1043 of the detected speech trigger. The weight adjustment unit 1043 learns the source data type of the speech trigger selected by the first user in the conversation with the second user based on the notification from the selection detection unit 1043. The weighting adjustment unit 1043 is based on the learning result in subsequent priority determination. As a result, a speech trigger presentation according to the conversation situation can be expected. Specifically, for example, even if the same speech trigger is presented to the first user, the priority may be higher or lower depending on the situation of the conversation.

  Hereinafter, the keyword conversion unit 1037 and the weight adjustment unit 1041 will be described in detail.

  FIG. 5 shows functional blocks of the keyword conversion unit 1037.

  The keyword conversion unit 1037 includes a strong conversion unit 601A, a weak change unit 601B, and a selection unit 603.

  Both the strong conversion unit 601A and the weak conversion unit 601B are modules that perform keyword conversion. The strength (complexity) of the conversion performed by the strong conversion unit 601A is stronger than the strength of the conversion performed by the weak conversion unit 601B. The stronger the conversion is, the more difficult it is to identify the keyword that is the base from the converted keywords. For example, if the keyword input to the keyword conversion unit 1037 is the keyword “married” after the change in the second user profile data, if converted by the conversion unit 601B having a low conversion strength, the relevance to marriage Keywords that are strong (keywords that are easily associated with “marriage”) are output. On the other hand, when converted by the conversion unit 601A with high conversion strength, keywords that are weakly related to marriage (the word “marriage” is directly Keywords that cannot be associated) are output.

  The selection unit 603 selects whether the keyword received by the keyword conversion unit 1037 is distributed to the strong conversion unit 601A or the weak conversion unit 601B. At that time, the closeness between the first and second users is referred to as appropriate. Specifically, for example, when the received keyword is a keyword from the second user profile data, the selection unit 603 determines that the closeness corresponding to the attribute item of the keyword is the parent between the first and second users. If it is lower than the density, the keyword is sent to the weak conversion unit 601B, and if the closeness is the same, the keyword is sent to the strong conversion unit 601A. For example, if the received keyword is a keyword from the utterance extraction unit 1035, the selection unit 603 sends the keyword to the strong conversion unit 601A.

  In this way, the keyword conversion strength can be varied based on the type of keyword and the familiarity between the first and second users.

  FIG. 6 shows functional blocks of the weight adjustment unit 1041.

  The weight adjustment unit 1041 manages the history data 650 for each user, and includes a history update unit 672 and an output control unit 671.

  In the history data 650, a history of transition of the source data type in the current (or so far) conversation between the first and second users is recorded. The “transition history of source data type” is a history of transition of the source data type of the keyword that is the base of the speech trigger selected by the first user. Source data types include profile data, conversation log data, and general data.

  The history update unit 672 receives the notification of the selected speech trigger from the selection detection unit 1043, and adds the source data type of the keyword that is the base of the speech trigger to the history data 650. The source data type is determined by the following method. In other words, for example, the output control unit 671 manages the source data type of the base keyword for each speech trigger output to the output unit 701. If the keyword is from the keyword extraction unit 1039, the source data type is managed as general data, and if the keyword is from the keyword conversion unit 1037, the source data type notified together with the converted keyword ( Profile data or conversation log data). The history update unit 672 can determine the source data type to be added by inquiring of the output control unit 671 the source data type corresponding to the speech trigger received from the selection detection unit 1043.

  The output control unit 671 creates the keyword itself from the keyword conversion unit 1037, the statement extraction unit 1035, and the keyword extraction unit 1039 or a question sentence generated based on the keyword as a statement trigger. The output control unit 671 determines the priority order of each speech trigger to be output based on the history data 650. Specifically, for example, according to the history data 650 illustrated in FIG. 6, the output control unit 671 indicates that the source data type has been “conversation log data” for a predetermined number of consecutive times. The priority of the speech trigger based on the extracted keyword is increased. This “raising” may be higher than the speech trigger corresponding to another source data type, or corresponding to the source data type “conversation log data” when output immediately before It may be higher than the priority level of the speech trigger.

  As mentioned above, although embodiment of this invention was described, this embodiment is only the illustration for description of this invention, and is not the meaning which limits the scope of the present invention only to this embodiment. The present invention can be implemented in various other modes without departing from the gist thereof.

  For example, in the above embodiment, the conversation is performed by inputting / outputting characters, but may be performed by inputting / outputting voice instead of or in addition thereto. Moreover, the said embodiment is applicable also to other kinds of conversations via a communication network, such as a video conference. The above embodiment can also be applied to a conversation in which three or more people participate (in this case, for example, the common term extraction unit 1033 is common to three or more users participating in one conversation. Information items).

  In addition, for example, instead of or in addition to the history data 650, other types of data that are updated each time a speech trigger is selected, such as data recorded for each source data type, the number of times the first user has been selected as the speech trigger. Data may be adopted.

  Further, the output control unit 671 may determine the display mode (for example, display position, character thickness, color, etc.) of the speech trigger instead of or in addition to the priority order of the speech trigger. Specifically, for example, the output control unit 671, when there is the largest amount of conversation log data as the source data type of the selected speech trigger, the display mode of the speech trigger based on the keyword from the conversation log data May be different from the display mode of the speech trigger corresponding to other source data types.

  Further, for example, the keyword may be acquired from the second user profile data regardless of the closeness between the first and second users. In this case, for example, the keyword conversion unit 1037 may control the conversion strength of the received keyword according to the security level of the attribute item corresponding to the keyword. For example, the selection unit 603 distributes keywords corresponding to items with a security level “high” (for example, [religion], [supporting political party]) to the strong conversion unit 600A, and sets an item with a security level “low” (for example, [birthplace ]) May be distributed to the weak conversion unit 600B.

  Further, the selection detection unit 1043 can detect a speech trigger input by another type of method instead of or in addition to the speech trigger selected on the window 950. For example, the selection detection unit 1043 compares a character string (for example, a character string input with a keyboard) input from the first user by a key operation on the message input screen with a message trigger presented in the window 950, When it is detected that the character string matches or is associated with any of the speech triggers presented in the window 950, the speech trigger can be notified to the weight adjustment unit 1041.

1 shows a computer system according to an embodiment of the present invention. It is explanatory drawing of the outline | summary of one Embodiment of this invention. An example of the change of the speech trigger according to an other party's speech is shown. The functional block of a conversation support part is shown. The functional block of a keyword conversion part is shown. The functional block of a weight adjustment part is shown.

Explanation of symbols

50 ... User terminal 100 ... Conversation space providing server

Claims (9)

An apparatus for supporting a conversation performed via a communication network between a first user using a first user terminal and a second user using a second user terminal,
Extracting means for extracting a character string from second user related data related to the second user;
Conversion means for converting the character string extracted by the extraction means into another character string related to the character string;
Creating means for creating a speech trigger that is a character string representing a trigger of speech for the second user of the first user, based on the character string of the result converted by the conversion means;
Transmission means for transmitting the created speech trigger to the first user terminal,
The conversion means performs a conversion of complexity corresponding to the familiarity representing the closeness between the first user and the second user on the character string extracted by the extraction means,
The intimacy is a value determined based on a conversation log corresponding to a conversation performed by the second user with the first user.
Conversation support device.   An apparatus for supporting a conversation performed via a communication network between a first user using a first user terminal and a second user using a second user terminal,
  Extraction means for extracting a character string from second user-related data including character strings corresponding to various items of the profile relating to the second user;
  Creation means for creating a speech trigger that is a character string representing a trigger of speech for the second user of the first user, based on the extracted character string or a character string after conversion of the character string;
  Transmitting means for transmitting the created speech trigger to the first user terminal;
With
  The extraction means extracts a character string corresponding to an item according to a closeness representing a closeness between the first user and the second user;
  The intimacy is a value determined based on a conversation log corresponding to a conversation performed by the second user with the first user.
Conversation support device. As the second user-related data, a second user profile data string is written corresponding to the various items, log conversations the second user is performed by have a conversation log data written,
The types of character strings extracted by the extracting means include the following (A) to (C):
The creation means creates one or more speech triggers based on the converted character strings of (A) and (B) and the character string of (C) that has not been converted,
The conversation support apparatus according to claim 2 .
(A) a character string extracted from the second user profile data;
(B) the second user is extracted from the conversational log corresponding to conversation conducted with the user other than the first user string,
(C) A character string extracted from the conversation log. The extraction means searches for data related to the converted character string from a site connected to the communication network, extracts the character string from the searched data,
The creation means further creates the speech trigger based on a character string extracted from the retrieved data;
The conversation support device according to any one of claims 1 to 3. An apparatus for supporting a conversation performed via a communication network between a first user using a first user terminal and a second user using a second user terminal,
An extraction means for extracting a character string from a plurality of different types of source data ;
Creating means for creating a speech trigger that is a character string representing a trigger of speech for the second user of the first user based on the extracted character string;
The display mode of the speech trigger is determined based on history data indicating the history of what kind of source data has been selected and the source data type of the character string that is the basis of the generated speech trigger. Display mode determining means;
Transmitting means for transmitting the speech trigger according to the determined display mode to the first user terminal ;
Selection detection means for detecting a speech trigger selected from the first user;
Source identifying means for identifying the type of source data of the character string that is the basis of the detected speech trigger;
A conversation support device comprising history update means for updating the history data based on the identified source data type .   The plurality of source data includes second user related data which is source data including character strings corresponding to various items of the profile relating to the second user;
  The extraction means extracts a character string corresponding to an item corresponding to an intimacy representing a closeness between the first user and the second user from the second user related data,
  The intimacy is a value determined based on a conversation log corresponding to a conversation performed by the second user with the first user.
The conversation support device according to claim 5. A method for supporting a conversation performed via a communication network between a first user using a first user terminal and a second user using a second user terminal,
A computer extracting a character string from second user related data relating to the second user;
A computer converting the character string extracted by the extracting means into another character string related to the character string;
The computer creates a speech trigger that is a character string representing a speech trigger for the second user of the first user based on the character string of the converted result;
Computer, possess and transmitting the speech trigger created in the first user terminal,
In the step of converting, the extracted character string is subjected to a conversion of complexity corresponding to the familiarity representing the closeness between the first user and the second user,
The intimacy is a value determined based on a conversation log corresponding to a conversation performed by the second user with the first user.
Conversation support method.   A method for supporting a conversation performed via a communication network between a first user using a first user terminal and a second user using a second user terminal,
  A computer extracting a character string from second user-related data including character strings corresponding to various items related to the second user;
  The computer creates a speech trigger that is a character string representing a trigger of speech for the second user of the first user, based on the extracted character string or a character string after conversion of the character string;
  A computer transmitting the created speech trigger to the first user terminal;
Have
  In the extracting step, a character string corresponding to an item corresponding to a closeness representing a closeness between the first user and the second user is extracted,
  The intimacy is a value determined based on a conversation log corresponding to a conversation performed by the second user with the first user.
Conversation support method.   A method for supporting a conversation performed via a communication network between a first user using a first user terminal and a second user using a second user terminal,
  A computer extracting a character string from a plurality of different types of source data including second user related data related to the second user;
  A computer creating a speech trigger based on the extracted character string, which is a character string representing a speech trigger for the second user of the first user;
  The display of the speech trigger based on the history data representing the history of what kind of speech trigger has been selected by the computer and the type of the source data of the character string on which the speech trigger was created. Determining an aspect;
  A computer transmitting the speech trigger according to the determined display mode to the first user terminal;
  Detecting the speech trigger selected from the first user;
  Identifying the type of character string source data that is the basis of the detected speech trigger;
  Updating the historical data based on the identified source data type;
A conversation support method.

Images