JP4388937B2 - Profile creation method and system - Google Patents

Description

  The present invention relates to a technology for supporting communication between users who transmit and receive text messages in real time on a conversation system.

  In the present invention, the conversation system is a system including a plurality of conversation devices. The conversation devices are connected to a network, can share a virtual space established on the network, and can send and receive text messages to each other in real time.

  IRC (Internet Relay Chat) is a type of conversation system constructed in accordance with the IRC protocol (RFC 1459). The IRC is configured by connecting an IRC server 11 and an IRC client 12 via the Internet. The IRC clients 12 share a virtual space called a channel with each other and send and receive text messages in real time. The IRC server 11 broadcasts a message from the IRC client 12 to other IRC clients 12 in the same channel. In the IRC, each IRC client 12 is uniquely specified by an identifier called a nickname. Each channel is given a channel name that uniquely identifies the channel.

  In recent years, a conversation system in which a plurality of users can talk in real time has been widely used as a means of communication. Specific examples of conversation systems include IRC and Nifty forums that are spreading due to the spread of the Internet. On the other hand, with the spread of the Internet, an access control service for web pages published by a WWW (World Wide Web) server and various web page information search services have been developed and provided. In providing these services, URLs (Uniform Resource Locators, RFC1738) have been classified according to predetermined categories. As a result, the URL is being used as an index indicating a category.

  On a conversation system such as IRC, there are tens of thousands of virtual spaces, and users over a wide physical range share the virtual spaces with each other. In entering such a virtual world and talking with other users, the user knows in advance what kind of person the other person sharing the virtual space is and what kind of conversation should be done. Is difficult. For example, in a conversation in real space, the content of the conversation and the timing of the conversation are coordinated with each other based on various information such as the relationship between you and the other party, the facial expression and clothes of the other party, and where the conversation is taking place. .

  However, it is difficult for the user to determine what kind of person the partner is and what to say in the virtual space in a conversation in a virtual space where it can be said that there is no information about the partner. As a result, the user cannot adjust the communication normally performed in the real space, the conversation content in the conversation space is boring, is spoken when trying to get out of the conversation space, or suddenly in a language other than the native language There are situations where people can speak or get uncomfortable remarks.

  The present invention provides a technique for facilitating communication on a network by enabling a user to provide preliminary knowledge about a conversation partner when starting a conversation in a conversation in a chaotic virtual space. Objective.

In order to solve the above-described problem, the first invention of the present application is a plurality of messages that are connected to a network and that can share one of a plurality of virtual conversation spaces established on the network and can transmit and receive messages to and from each other. In a conversation system including a plurality of conversation devices and a conversation space generation device for generating the conversation space, the profile creation method executed by the conversation device or the conversation space generation device,
A predetermined keyword and a predetermined category are associated with each other and stored in the category storage unit,
A message that is transmitted / received in the virtual space is acquired from a conversation system, and when the keyword is included in the acquired message, a category corresponding to the keyword is specified based on the category storage means, and the message includes the keyword Further speaking time from the conversation system, write the keyword speaker, the keyword and the identified category in association with the user information storage means,
Create a conversation history in the conversation space, and at least a predetermined number n of messages (n is a positive integer) uttered before the keyword utterance or a predetermined number m (m is a positive integer) uttered after the keyword utterance Among the messages, for a message that is said within a predetermined time interval from the keyword's statement , the speaker of the message is acquired from the conversation history, and the user information is associated with the keyword as a related speaker. Write more to the storage means,
Provide a profile creation method.

  A case where a URL is used as a keyword will be described as an example. A category table in which URLs and predetermined categories are associated with each other is prepared in advance. An existing DB can be used as the category table. It is determined whether a URL is included every time a message is made. If the URL is included, the category corresponding to the URL is determined based on the category table. Next, the user who made the URL and the determined category are stored in association with each other. Users can be classified into categories based on their own remarks.

  Messages spoken before and after the keyword utterance are considered to be closely related to the keyword utterance. Therefore, various services can be provided by storing the speakers of these messages in the user table. For example, by specifying a keyword and a speaker, it is possible to participate in a virtual space where the user has participated before.

  In addition, the time when the user speaks the keyword is stored in the user table. The keyword valid time can be further stored in the user table. The valid time may be stored in the management unit in advance, or the management unit may accept the setting of the valid time from the user. By deleting an entry for a keyword whose expiration date has passed from the user table, a user profile reflecting the actual user state can be created.

A second invention of the present application includes a plurality of conversation devices that are connected to a network and share any one of a plurality of virtual conversation spaces set up on the network and are capable of transmitting and receiving messages to each other, and the conversation space. In a conversation system including a conversation space generation device to be generated, a profile creation method executed by the conversation device or the conversation space generation device is provided. The method includes the following steps.
A predetermined keyword described in a predetermined format and a predetermined category are associated with each other and stored in the category storage means;
The user, the predetermined keyword that the user has spoken, the category, and the utterance time of the keyword are associated and stored in the user information storage unit,
The message sent to the conversation space is acquired from the conversation system, and it is determined whether or not the acquired message includes a keyword described in a predetermined format, and the keyword described in the predetermined format is included. A category corresponding to the keyword described in the predetermined format is determined on the basis of the category storage means, further obtaining a utterance time of a message including the keyword from the conversation system, and described in the predetermined format. The keyword, the keyword, the determined category, and the speech time are written in the user information storage means,
A conversation history is created in a conversation space, and at least a predetermined number n (n is a positive integer) of messages uttered before a keyword utterance or at least a predetermined number m (m is a positive integer) uttered after a keyword utterance Among the messages of the above, for a message that is said within a predetermined time interval from the keyword's statement, the speaker of the message is acquired from the conversation history, and the user is associated with the keyword as a related speaker Further writing to the information storage means.

  The same effects as the first invention are achieved.

The third invention of the present application includes a plurality of conversation devices that are connected to a network and that can share one of a plurality of virtual conversation spaces established on the network and can send and receive messages to each other. A profile creation system used in a conversation system is provided. This system has the following means.
Category storage means for storing predetermined keywords and predetermined categories in association with each other;
User information storage means for storing the user, the predetermined keyword said by the user, the category, and the keyword utterance time in association with each other;
A message sent to the conversation space is acquired from the conversation system, and when the keyword is included in the message, a category corresponding to the keyword is determined based on the category storage means, and the message including the keyword is remarked Management means for further acquiring the time from the conversation system and writing the speaker of the keyword, the keyword, the determined category, and the speech time in the user information storage means.

The management means creates a conversation history in a conversation space, and at least a predetermined number n (n is a positive integer) of messages uttered before a keyword is uttered or at least a predetermined number m (m is a utterance after a keyword is uttered) Among messages of a positive integer), for a message that is spoken within a predetermined time interval from the keyword, the message speaker is obtained from the conversation history, and the corresponding speaker is associated with the keyword. In addition, the information is further written in the user information storage means.

  The same effects as the first invention are achieved.

The fourth invention of the present application includes a plurality of conversation devices connected to a network and capable of transmitting and receiving messages to each other by sharing any of a plurality of virtual conversation spaces established on the network. A profile creation system used in a conversation system is provided. This system has the following means.
Category storage means for storing a predetermined keyword described in a predetermined format and a predetermined category in association with each other;
User information storage means for storing the user, the predetermined keyword said by the user, the category, and the keyword utterance time in association with each other;
The message sent to the conversation space is acquired from the conversation system, and it is determined whether or not the acquired message includes a keyword described in a predetermined format, and the keyword described in the predetermined format is included. A category corresponding to the keyword described in the predetermined format is determined on the basis of the category storage means, further obtaining a utterance time of a message including the keyword from the conversation system, and described in the predetermined format. Management means for writing the keyword speaker, the keyword, the determined category, and the speech time in the user information storage means.

The management means creates a conversation history in a conversation space, and at least a predetermined number n (n is a positive integer) of messages uttered before a keyword is uttered or at least a predetermined number m (m is a utterance after a keyword is uttered) Among messages of a positive integer), for a message that is spoken within a predetermined time interval from the keyword, the message speaker is obtained from the conversation history, and the corresponding speaker is associated with the keyword. In addition, the information is further written in the user information storage means.

  The same effects as the first invention are achieved.

The fifth invention of the present application is the third or fourth invention, wherein
The user information storage means further stores, for each user, rating information indicating the degree of harmfulness of the user in addition to the user, the keyword and the category;
The management unit provides a profile creation system that updates user rating information when a predetermined harmful keyword is spoken.

  For example, it is conceivable that the initial value of the rating information of the user is set to “0” and the value of the rating information is increased every time the user utters a harmful keyword.

  Depending on the value of the rating information, it becomes possible to refuse the participation in the channel or not to broadcast the message in the channel, and the safety of the virtual community can be increased.

  A sixth invention of the present application provides the profile creation system according to the second invention, wherein the conversation system processes a message transmitted and received by the conversation device based on a user category stored in the user table.

  For example, take IRC as an example. When the IRC client 12 specifies a category together with channel creation, the IRC server 11 acquires users belonging to the specified category from the management unit, and notifies only the acquired user of channel creation. It is also possible to manage messages based on the rating information.

The seventh invention of the present application is used in a conversation system including a plurality of conversation devices that are connected to a network and share any one of a plurality of virtual conversation spaces established on the network and can transmit and receive messages to each other. A computer-readable recording medium storing a profile creation program to be recorded is provided. This recording medium stores a profile creation program for executing the following steps.
A; a step of storing a predetermined keyword and a predetermined category in association with each other in the category storage means;
B; obtaining a message sent to the conversation space from the conversation system;
C; when a keyword is included in the message, determining a category corresponding to the keyword based on the category storage means;
D: a step of further acquiring a utterance time of a message including a keyword from the conversation system and storing the utterer of the keyword, the keyword, the determined category, and the utterance time in association with each other in the user information storage unit ,
E: Create a conversation history in the conversation space, and at least a predetermined number n (n is a positive integer) of messages uttered before the keyword is uttered, or at least a predetermined number m (m is a positive integer) uttered after the keyword is uttered ) For messages that have been made within a predetermined time interval from the keyword's utterance, the speaker of the message is obtained from the conversation history, and is associated with the keyword as a relevant speaker. Further storing in the user information storage means.

  It has the same effect as the first invention.

The eighth invention of the present application is used in a conversation system including a plurality of conversation devices that are connected to a network and share any one of a plurality of virtual conversation spaces established on the network and can transmit and receive messages to each other. A computer-readable recording medium storing a profile creation program to be recorded is provided. This recording medium stores a profile creation program for executing the following steps.
A step of storing a predetermined keyword described in a predetermined format and a predetermined category in association with each other in the category storage means;
A step of storing the user, the predetermined keyword said by the user, the category, and the utterance time of the keyword in association with each other in the user information storage unit;
The message sent to the conversation space is acquired from the conversation system, and it is determined whether or not the acquired message includes a keyword described in a predetermined format, and the keyword described in the predetermined format is included. A category corresponding to the keyword described in the predetermined format is determined on the basis of the category storage means, further obtaining a utterance time of a message including the keyword from the conversation system, and described in the predetermined format. Writing the keyword's speaker, the keyword, the determined category, and the speech time in the user information storage means;
A conversation history is created in a conversation space, and at least a predetermined number n (n is a positive integer) of messages uttered before a keyword utterance or at least a predetermined number m (m is a positive integer) uttered after a keyword utterance Among the messages of the above, for a message that is said within a predetermined time interval from the keyword's statement, the speaker of the message is acquired from the conversation history, and the user is associated with the keyword as a related speaker Further writing to the information storage means.

  The same effects as the first invention are achieved.

  By using the present invention, it is possible to create a profile that grasps the user's interests and characteristics in real time. By using this profile, information that helps communication in the virtual space can be provided to the user, and smooth communication can be promoted.

  Next, the profile creation system of the present invention will be specifically described with reference to an embodiment.

<First embodiment>
In this embodiment, a case where the profile creation system is applied to IRC will be described as an example. However, this system can be similarly applied to other conversation systems.

[Constitution]
FIG. 1 is an overall configuration diagram of a profile creation system 1 according to the first embodiment. The profile creation system 1 according to the present embodiment includes a category DB 13, a user information DB 14, a user information management server 15, a message management unit 16, and a rating processing unit 17 in addition to the IRC server 11 and the IRC client 12 that constitute the IRC. It consists of

  In the present embodiment, the speech management unit 16 and the rating processing unit 17 are provided on the IRC server 11 side, but a configuration may be provided on the IRC client 12 side. The rating processing unit 17 is a constituent element necessary when the user information DB 14 has rating information described later.

(1) Category DB
In the category DB 13, predetermined keywords and predetermined categories are associated and stored. In this embodiment, a case where a URL is used as a keyword will be described as an example. FIG. 2 is a conceptual explanatory diagram of a category table stored in the category DB 13. The category table stores, for each URL, which category each URL belongs to. For example, “URL1” belongs to the category “hobby”. One URL may belong to a plurality of categories.

  As the URL category table, for example, data included in a search engine operating on the Internet, URL filtering software data, or the like can be used. It is also possible to prepare and use a category table in which predetermined words such as “car” and “travel” are associated with a category without using a URL as a keyword.

(2) User information DB
Although it does not specifically limit as user information DB14, For example, the database which supports LDAP (Light Weight Directory Access Protocol) which is a protocol for user information databases is used preferably. FIG. 3 is a conceptual explanatory diagram of a user information table stored in the user information DB 14. In the user information table, user information is stored for each user. In the present embodiment, the user information includes a user ID, a password, chat related information, personal information, and speech attribute information.

  The user ID is identification information given to each user in order to uniquely identify the user. FIG. 3 shows user information identified by the user ID “ID0001”.

  The password is authentication information used when updating user information.

  In the chat related information, a user's nickname, a currently participating channel, a client host name that uniquely identifies the user terminal, a server name to which the user terminal is connected, and the like are described.

  Personal information includes information such as name, age, and occupation. Generally, personal information is static information that is input and registered by the user. Moreover, rating information can also be included in personal information. The rating information is information indicating that the higher the value is, the more harmful the user is. Rating information will be described later.

  In the comment attribute information, URLs spoken by the user are described for each category defined in the category DB 13. Specifically, the speech attribute information includes categories, the number of speeches included in each category, and speech information. In the speech information, the speech number, speech content, channel, date, time limit, and related speaker are described.

  As the “category”, any category defined in the category DB 13 is described. In FIG. 3, “hobby” and “Internet” are described as categories.

  “Number of utterances” indicates the number of utterances of URLs associated with the category. For example, in FIG. 3, 15 URLs are associated with the category “hobby”.

  The “speech number” is an identification number that identifies a speech specified by the speech information within one category. For example, in FIG. 3, “utterance 1”, “utterance 2”, and the like correspond to the speech numbers.

  In the “utterance content”, a keyword that is spoken by the user, that is, a URL is described. In FIG. 3, the user “ID0001” speaks “URL1”, “URL2”, and the like as URLs belonging to the category “hobby”.

  “Channel” describes the name of the channel in which the URL is spoken. For example, the “URL1” is stated on channel #AAA.

  “Date” describes the date and time when the URL was spoken. The reason is to set the expiration date. For example, in FIG. 3, URL1 is remarked 1999/6/7, 13:55.

  “Time limit” describes the length of time for which the statement specified by the statement number is valid. For example, “Remark 1” in the category “Hobby” in FIG. 3 is valid for one month from 13:55 on June 7, 1999.

  The “related speaker” describes the ID of the user who is speaking another message before and after the message including the URL is spoken. For example, FIG. 3 shows that users “ID0002” and “ID0003” are speaking before and after URL1. In the present embodiment, a user who speaks the message immediately before the URL and the two messages immediately after the URL is a related speaker.

(3) Rating information Rating information is information indicating the degree of harmfulness of a user. In the present embodiment example, the harmful speech level is used as the rating information. The rating information, that is, the harmful speech level is initially set to “0”, for example. Each time a predetermined URL stored in advance as a harmful URL is uttered, the degree of harmful utterance of the speaker increases.

(4) User information management server 15 and speech management unit 16
The user information management server 15 registers and updates user information in the user information DB 14.

  The message management unit 16 acquires a message in each channel from the IRC server, and performs processing to update the message attribute information in the user information DB 14. Further, the comment management unit 16 notifies the rating processing unit 17 of the update of the comment attribute information. Further, the comment management unit 16 can perform a predetermined process based on the comment attribute information. For example, the comment management unit 16 notifies the IRC server 11 of other users belonging to a certain category, or sends the comment attribute information to the IRC server 11. Note that the message management unit 16 creates and stores a message history for each channel.

(5) Rating processing unit 17
The rating processing unit 17 updates the rating information of the user whose message attribute information is updated in accordance with the notification from the message management unit 16. Specifically, URLs that are contrary to public order and morals are stored in the rating processing unit 17 as harmful URLs in advance. The rating processing unit 17 calculates the rating information of the speaker when a new URL is spoken. The rating processing unit 17 instructs the user information management server 15 to rewrite the rating information. Furthermore, the rating processing unit 17 can determine how to process a user's utterance based on the rating information in the user information table, and can instruct the IRC server 11.

  The rating information is calculated as follows, for example. The rating processing unit 17 stores a harmful attribute list shown in FIG. FIG. 4 is a conceptual explanatory diagram of a harmful attribute list held by the rating processing unit 17. In the noxious attribute list, noxiousness weights C1, C2, C3... Are stored for each category. The rating processing unit 17 calculates the harmful speech level Xn of each user using the weight of each category.

  Specifically, the harmful speech level Xn is obtained as the sum of the weights of each category multiplied by the number of speeches in the corresponding category of the user. For example, the harmful speech level Xn of the user “ID0001” having a harmful speech count of 15 in the category “hobby” and a harmful speech count of 30 in the “Internet” is calculated as 15 × C1 + 30 × C3 = Xn.

  Furthermore, the rating processing unit 17 can also perform predetermined processing using the rating information. FIG. 5 is a conceptual explanatory diagram of a processing table held by the rating processing unit 17. The rating processing unit 17 determines processing based on the rating information and instructs the IRC server 11.

  In the example of FIG. 5, if the harmful speech level Xn ≧ X1, the rating processing unit 17 instructs the server to disconnect the connection with the user terminal. If X1> Xn ≧ X2, the message from the user is canceled on the server side and is not broadcast to other users. If X2> Xn ≧ X3, a speech from the user is not accepted until a predetermined time has elapsed. If X3> Xn ≧ X4, when the corresponding user connects to the IRC server 11, a warning is transmitted from the server to the corresponding user.

[Process flow]
(1) Statement Attribute Update Processing FIG. 6 is a flowchart showing the flow of statement attribute update processing performed by the statement management unit 16. For ease of explanation, a case where the message shown in FIG. 7 is made on channel #AAA is taken as an example. Further, it is assumed that the user ID of the user “taro” who made the URL is “ID0001” and the user information is shown in FIG. As described above, what is written in the comment attribute information as related speakers is the speaker “hanako” of the message immediately before the URL, and the speakers “jiro” and “saburo” of the two messages immediately after the URL. When a message is spoken in any channel, the following processing is started.

  In step S1, the message management unit 16 acquires the generated message from the IRC server 11 and stores it in the message history for each channel.

  In step S <b> 2, the speech management unit 16 acquires necessary information for specifying the user from the IRC server 11. For example, a nickname, an IP address, a client host name, a login name, and the like can be given. In this embodiment, the client host name is acquired from the IRC server 11.

  In step S3, the speech management unit 16 determines whether or not a URL has been spoken in the past for all users in the channel #AAA where the speech has occurred. The utterance management unit 16 determines whether or not it is necessary to write a speaker of a newly uttered URL as a related speaker for a user who has uttered a URL in the past. For this determination, the speech management unit 16 uses the monitoring target list created in step S9 described later.

  FIG. 8 is a conceptual explanatory diagram of the monitoring target list. The monitoring target list is created for each user. For example, as shown in FIG. 7, when the user “taro” utters URL1, the following contents are written in the monitoring target list of the user “taro”.

Comment content column; “URL1”
Channel field: Channel name "#AAA" of the channel where the URL was spoken
Date column; date and time when a message including a URL occurred Number of monitoring utterances: “2”.

  For each item of message content, channel, and date, since the same content is registered in the message attribute information of the user information (FIG. 3), only the address of the corresponding item in the user information is stored, It is also possible to acquire the contents of actual data from user information.

  The number of monitoring utterances indicates the number of messages regarded as messages related to the URL among messages made immediately after the message including the URL (hereinafter referred to as the URL). In the illustrated example, the number of monitoring utterances is set for a message uttered immediately after the URL, but the number of monitoring utterances may be set immediately before and after. Also, the number of messages considered to be related to the URL need not be uniform before and after the URL. For example, the number of messages considered to be related may be set separately immediately before and after the utterance of the URL, with one message immediately before the URL and five messages immediately after the URL.

  Also, the number of monitored utterances decreases by one each time an utterance immediately after the message is detected. When the number of monitoring utterances reaches 0, the corresponding entry is deleted from the monitoring list. In addition, a monitoring message count area may be provided separately from the number of monitoring messages, and the operation may be determined according to the value of the count area.

  The message management unit 16 determines whether there is a monitoring target list regarding the channel #AAA where the message is generated. When the monitoring target list exists, the process proceeds to step S12 described later. If not, the process proceeds to step S4.

  In step S4, the comment management unit 16 searches the category table. Specifically, it is determined whether or not a URL is included in the user's utterance. If the URL is included, the category table is searched by using the URL as a key to determine the category. In this example, since URL1 is spoken, the category is determined as “hobby”.

  In step S <b> 5, the comment management unit 16 determines whether there is a category for the spoken URL or whether the category is not included. If there is a category, the process proceeds to step S6. If there is no category corresponding to the replied URL or the URL is not included, the process proceeds to step S10 described later.

  In step S <b> 6, the message management unit 16 acquires a predetermined number of messages remarked immediately before the URL. As described above, when the number of messages to be spoken immediately before the URL is set in the monitoring target list, the speech management unit 16 sets the number of messages to be acquired by referring to the monitoring target list. decide. The message management unit 16 acquires these messages from the message history accumulated for each channel. The utterance history stores the nickname of the speaker of the message along with the message. The utterance management unit 16 acquires the nickname of the speaker of the message, in this example “hanako”, from the utterance history as the nickname of the related speaker. Furthermore, the comment management unit 16 refers to the user information DB 14 and acquires a user ID for the acquired nickname “hanako”, for example, “ID0002”.

  In step S7, the comment management unit 16 acquires the comment time of the commented URL from the comment history. Moreover, the expiration date of this statement is set. The reason for setting the expiration date is to delete the message from the message attribute information because it is considered that the user's current state is not properly reflected in the case of a message that is too old. As the expiration date, for example, the same expiration date may be set for all statements. Different expiration dates may be set for each category. It is also possible to accept an expiration date setting from the user.

  In step S8, the speech management unit 16 sets the spoken URL 1, the related speaker acquired in step S6, the speech time and the expiration date set in step S7 to the category determined in step S4 of the speech attribute information of the corresponding user. to add. In this example, the message number “utterance 16” is added to the category “hobby”. In addition, the comment management unit 16 updates the number of comments in the category to which the comment attribute information is added. In this example, the number of utterances of “hobby” of the user “ID0001” is updated to “16”.

  In step S9, the comment management unit 16 provides an entry for the URL1 in the monitoring target list of the user specified in step S2. In other words, an entry corresponding to the utterance of the URL is provided in the monitoring target list of the utterer of the URL. FIG. 8 is a monitoring target list at this stage.

  In step S <b> 10, the speech management unit 16 determines whether a speech whose expiration date has passed is in the speech attribute information. If "Yes" is determined, the process proceeds to step S11. If "No" is determined, the process ends.

  In step S11, the comment management unit 16 deletes the entry of the comment attribute information that has already expired from the comment attribute information of the user information.

  If it is determined in step S3 that there is an entry in the monitoring target list, the process proceeds to step S12. In step S12, the speech management unit 16 writes the user ID of the speaker of the generated message in the speech attribute information of the specific user. The specific user is a user in the channel #AAA that made a statement and has a statement entry in the channel #AAA in the monitoring target list. Specifically, when “jiro” or “saburo” speaks after “taro”, it is written in the speech attribute information of “taro” as the related speaker of “utterance 16”.

  Further, the comment management unit 16 decrements the number of monitored comments in the monitoring target list. In this example, the number of monitored utterances specified by the utterance content “URL1”, channel “#AAA”, and date “1999/06/07 18:55” is subtracted by one.

  In step S13, the comment management unit 16 determines whether there is an entry to be deleted from the monitoring target list. Specifically, it is determined whether or not there is an entry whose monitoring message count is “0”. If "Yes" is determined, the process proceeds to step S14. If "No" is determined, the process proceeds to step S5.

  In step S14, the message management unit 16 deletes the entry whose monitoring message count is “0” from the monitoring target list. Thereafter, the process returns to step S4.

  Note that the speech management unit 16 may calculate a time interval between a message speech considered to be related and a speech of the URL. When this time interval exceeds a predetermined time set in advance, the speech management unit 16 may not write the acquired user ID of the related speaker in the speech attribute information. This is because when the time interval between both statements is long, the relevance of both statements is considered to be weak.

(2) Rating Information Update Processing FIG. 9 is a flowchart showing the flow of the rating information update processing by the rating processing unit 17 and the processing based on the rating information. When the comment management unit 16 is notified that the comment attribute information has been updated, the rating processing unit 17 starts the following processing.

  In step S21, the rating processing unit 17 acquires the user information table of the user whose statement attribute information has been updated. The user information table may be read from the user information management server 15 or acquired from the message management unit 16.

  In step S22, the rating processing unit 17 calculates the number of harmful utterances of the corresponding user for each category. Specifically, harmful URLs are detected from URLs spoken by the user, and how many times harmful URLs are spoken in each category is calculated.

  In step S23, the rating processing unit 17 calculates the harmful speech level Xn for the corresponding user.

  In step S24, the rating processing unit 17 determines a process according to the obtained harmful speech degree Xn based on the processing table, and instructs the IRC server 11. Thereafter, the process ends.

  Note that it may be determined first whether or not a newly-sent URL is harmful, and if it is harmful, the harmful speech level Xn may be updated.

[Example of services based on user information]
By using the comment attribute information and rating information created by the profile creation system, it is possible to provide various services for facilitating communication on a conversation system such as IRC. Specific examples of services are given below.

(1) Provision of service using speech attribute information (1-1) Talk with people belonging to the same category.

  For example, the user can invite only users belonging to a specific category to the channel by specifying a channel name and a category. For example, a command “INVITEEX #Channel Category” is prepared in advance in the IRC. Here, in “#Channel”, the name of the invitation destination channel is described. In “Category”, a category specified by the user is described. When this command is remarked from the IRC client 12, the IRC server 11 distributes the command to a user terminal belonging to the specified category. The message management unit 16 detects the command, searches for users belonging to the category specified by the command based on the user information DB 14, and notifies the IRC server 11 of the user.

  (1-2) Send a message to a person belonging to the same category.

  For example, a user can send a message to a user belonging to a specific category by specifying a channel name and a category. For example, a command “PRIVMSGEX Category MSG” is prepared in advance in the IRC. Here, “Category” describes a category designated by the user. “MSG” describes a text message input by the user. When this command is remarked from the IRC client 12, the IRC server 11 distributes the command to a user terminal belonging to the specified category. As in the case (1), the message management unit 16 detects a command, searches the user information DB 14, and notifies the IRC server 11 of the searched user.

(1-3) Creation of category-specific friend list For example, the user can display a category-specific friend list by the IRC client 12. In the IRC, the IRC server 11 notifies the IRC client 12 of predetermined information such as the nickname of the user in the channel. Therefore, the IRC server 11 may notify the user category in addition to the predetermined information. The IRC server 11 can acquire the user category from the statement management unit 16. The acquisition timing can be, for example, when an IRC client enters the channel.

  (1-4) Display the users in the channel by category.

  FIG. 10 is an example in which users in the channel are displayed by category. As described in (3) above, when the user's category is notified to the IRC client 12, the IRC client 12 can display the message from the user or the user by changing the color or font for each category. is there.

(2) Service based on rating information Communication can be facilitated based on rating information as in the case of speech attribute information. Examples other than those listed in the processing table of FIG. 5 are listed below.

  (2-1) Restrict channel creation and entry.

  The IRC server 11 can restrict channel creation and room entry for users with high rating information. For example, “prohibition of channel creation” and “prohibition of channel entry” are provided in the processing table of FIG. When the “JOIN #Channel” command is transmitted from the IRC client 12, the rating processing unit 17 determines whether to allow creation of the designated channel and entry into the channel based on the rating information of the sender and the processing table. Judge whether or not. The IRC server 11 permits or rejects entry into the channel according to the determination result.

  The “JOIN #Channel” command is an existing command on the IRC for requesting creation of a channel designated by “#Channel” and entry into the channel.

  (2-2) Limit the number that can participate in the channel.

  The IRC server 11 can limit the number that can participate in the channel according to the user's rating information. In this case, the maximum number of channels that can participate in each level of harmful speech level is set in the processing table of FIG. The rating processing unit 17 notifies the IRC server 11 of the number of channels that can participate in the user who has transmitted the JOIN command. Since the IRC server 11 knows the number of channels that each user is participating in, the IRC server 11 compares the number of channels that can be participated with the number of participations, and permits or rejects entry into the channel.

  (2-3) Blocking channel invitations from harmful users.

  For example, even if an invitation command to a certain channel is transmitted from a user whose rating information is a predetermined value or more, the IRC server 11 does not broadcast to other users. Similarly to the above, it is set in the processing table that the invitation command is ignored.

  In this case, the rating processing unit 17 determines processing for ignoring the invitation command from the rating information of the transmission source user, and notifies the IRC server 11 of the processing. Thereby, it can prevent being invited to a channel by a harmful user.

  (2-4) Preventing harmful users from being invited to the channel.

  In the processing table, it is set that a user whose rating information is a predetermined value or more is not transmitted to another user's channel invitation command.

  In this case, the rating processing unit 17 determines not to transmit the invitation command based on the rating information of the destination user, and notifies the IRC server 11 of it. The IRC server 11 stops the transmission of the invitation command to the harmful user according to the determination of the rating processing unit 17. Thereby, it is possible to prevent a harmful user from being invited to the channel without knowing it.

  (2-5) Restrict the speech itself.

  It is set in the processing table that character messages from users whose rating information is greater than or equal to a predetermined value are not distributed to other users. In addition, it is set that messages from other users are not distributed to the user.

  Based on the rating information of the speaker or the receiver, the rating processing unit 17 determines whether or not to distribute the speech and the delivery destination of the speech, and notifies the IRC server 11 of it. Thereby, inappropriate statements can be suppressed in advance. Further, it is possible to prevent other users' utterances from leaking to users who have many inappropriate utterances.

  (2-6) Control the operation of the agent.

  In a chat system, there is a case where a software robot, that is, a so-called agent, that creates a conversation history or responds in response to a user's remarks is operating. When a utterance or command is transmitted from the user whose rating information is a predetermined value or more to the agent, the rating processing unit 17 instructs the IRC server 11 to stop delivery of the utterance or command. As a result, user behavior that may cause the agent to operate inappropriately can be suppressed, and the safety of the community in the channel can be improved.

<Second Embodiment>
In the first embodiment, the example in which the user category is specified from the predetermined keyword in the text message has been described. However, it is also possible to specify a category using a keyword in a structured document such as XML or SGML as well as simple text. Further, a part other than the content in the structured document, that is, a so-called meta information part can be used as a basis for category specification.

  When specifying a category from a structured document, first, the structure analysis of the structured document is performed. Next, a category is specified for each tag, and the user's speech attribute information is updated in the same manner as in the first embodiment.

<Third Embodiment>
Similarly, in one-to-one conversation communication, it is possible to provide a service based on speech attribute information and rating information. Hereinafter, a specific example will be described.

  (1) Limit the conversation itself.

  It is possible to restrict the transmission of a conversation request according to the rating information of a user who desires a one-to-one conversation or a user who talks to. It is possible to prevent opening a one-to-one connection without knowing that the user is harmful.

  (2) Restrict acquisition of user information.

  The user information to be provided is limited according to the rating information of the user who wants to acquire the user information of other users. For example, in the processing table, it is conceivable to set user information to be disclosed step by step for each harmful speech level.

  (3) Restrict the speech itself.

  Similar to the first embodiment (2-5), it is possible to restrict the transmission of the speech according to the rating information of the speaker or the recipient.

1 is an overall configuration diagram of a profile creation system. The conceptual explanatory view of a category table. A conceptual explanatory view of a user information table. The flowchart which shows the flow of remark attribute update processing. An example of a message. The conceptual explanatory view of a monitoring object list. The conceptual explanatory view of a harmful attribute list. The conceptual explanatory view of a processing table. The flowchart which shows the flow of the rating information update process. Explanatory drawing which shows the specific example of the display according to category.

  1: Profile creation system

Claims (8)

A plurality of conversation devices that are connected to a network and share any one of a plurality of virtual conversation spaces established on the network and can send and receive messages to each other, and a conversation space that generates the conversation space In a conversation system including a generation device, a profile creation method executed by the conversation device or the conversation space generation device,
A predetermined keyword and a predetermined category are associated with each other and stored in the category storage unit,
A message that is transmitted / received in the virtual space is acquired from a conversation system, and when the keyword is included in the acquired message, a category corresponding to the keyword is specified based on the category storage means, and the message includes the keyword Further speaking time from the conversation system, write the keyword speaker, the keyword and the identified category in association with the user information storage means,
Create a conversation history in the conversation space, and at least a predetermined number n of messages (n is a positive integer) uttered before the keyword utterance or a predetermined number m (m is a positive integer) uttered after the keyword utterance Among the messages, for a message that is said within a predetermined time interval from the keyword's statement , the speaker of the message is acquired from the conversation history, and the user information is associated with the keyword as a related speaker. Write more to the storage means,
Profile creation method. A plurality of conversation devices that are connected to a network and share any one of a plurality of virtual conversation spaces established on the network and can send and receive messages to each other, and a conversation space generation device that generates the conversation space A profile creation method executed by the conversation device or the conversation space generation device,
A predetermined keyword described in a predetermined format and a predetermined category are associated with each other and stored in the category storage unit.
The user, the predetermined keyword that the user has spoken, the category, and the utterance time of the keyword are associated and stored in the user information storage unit,
A message sent to the conversation space is acquired from the conversation system, and it is determined whether or not a keyword described in a predetermined format is included in the acquired message, and a keyword described in a predetermined format is included. A category corresponding to the keyword described in the predetermined format is determined on the basis of the category storage means, further obtaining a speech time of a message including the keyword from the conversation system, and describing in the predetermined format The keyword speaker, the keyword, the determined category, and the speech time are written in the user information storage means,
Create a conversation history in the conversation space, and at least a predetermined number n of messages (n is a positive integer) uttered before the keyword utterance or a predetermined number m (m is a positive integer) uttered after the keyword utterance Among the messages, for a message that is said within a predetermined time interval from the keyword's statement , the speaker of the message is acquired from the conversation history, and the user information is associated with the keyword as a related speaker. Write more to the storage means,
Profile creation method. Profile used in a conversation system that is connected to a network and includes a plurality of conversation devices that share one of a plurality of virtual conversation spaces established on the network and can send and receive messages to each other. A creation system,
Category storage means for storing a predetermined keyword and a predetermined category in association with each other;
User information storage means for storing a user, a predetermined keyword spoken by the user, the category, and a speech time of the keyword in association with each other;
When a message sent to the conversation space is acquired from the conversation system and the keyword includes the keyword, a category corresponding to the keyword is determined based on the category storage means, and the message utterance time including the keyword And a management unit that writes the speaker of the keyword, the keyword, the determined category, and the utterance time to the user information storage unit,
The management means creates a conversation history in a conversation space, and at least a predetermined number n (n is a positive integer) of messages uttered before a keyword is uttered or at least a predetermined number m (m is a utterance after a keyword is uttered) Among messages of a positive integer), for a message that is spoken within a predetermined time interval from the keyword, the message speaker is obtained from the conversation history, and the corresponding speaker is associated with the keyword. And further writing to the user information storage means,
Profile creation system. Profile used in a conversation system that is connected to a network and includes a plurality of conversation devices that share one of a plurality of virtual conversation spaces established on the network and can send and receive messages to each other. A creation system,
Category storage means for storing a predetermined keyword described in a predetermined format and a predetermined category in association with each other;
User information storage means for storing a user, a predetermined keyword spoken by the user, the category, and a speech time of the keyword in association with each other;
A message sent to the conversation space is acquired from the conversation system, and it is determined whether or not a keyword described in a predetermined format is included in the acquired message, and a keyword described in a predetermined format is included. A category corresponding to the keyword described in the predetermined format is determined on the basis of the category storage means, further obtaining a speech time of a message including the keyword from the conversation system, and describing in the predetermined format Management means for writing a keyword speaker, a keyword, the determined category, and the speech time in the user information storage means,
The management means creates a conversation history in a conversation space, and at least a predetermined number n (n is a positive integer) of messages uttered before a keyword is uttered or at least a predetermined number m (m is a utterance after a keyword is uttered) Among messages of a positive integer), for a message that is spoken within a predetermined time interval from the keyword, the message speaker is obtained from the conversation history, and the corresponding speaker is associated with the keyword. And further writing to the user information storage means,
Profile creation system. In addition to the user, the keyword and the category, the user information storage means further stores rating information indicating the degree of harmfulness of the user for each user,
The profile creation system according to claim 3 or 4, wherein the management unit updates the user rating information when a predetermined harmful keyword is spoken.   The profile creation system according to claim 2, wherein the conversation system processes a message transmitted and received by the conversation apparatus based on a user category stored in the user table. A profile creation program used in a conversation system including a plurality of conversation devices connected to a network and sharing one of a plurality of virtual conversation spaces established on the network and capable of transmitting and receiving messages to each other is stored. A computer-readable recording medium,
A; associating a predetermined keyword with a predetermined category and storing them in the category storage means;
B; obtaining a message sent to the conversation space from the conversation system;
C; when a keyword is included in the message, determining a category corresponding to the keyword based on the category storage means;
D: a step of further acquiring a utterance time of a message including a keyword from the conversation system and storing the utterer of the keyword, the keyword, the determined category, and the utterance time in association with each other in the user information storage unit When,
E: Create a conversation history in the conversation space, and at least a predetermined number n (n is a positive integer) of messages uttered before the keyword is uttered, or at least a predetermined number m (m is a positive integer) uttered after the keyword is uttered ) For messages that have been made within a predetermined time interval from the keyword's utterance, the speaker of the message is obtained from the conversation history, and is associated with the keyword as a relevant speaker. Further storing in the user information storage means;
The computer-readable recording medium which memorize | stored the profile creation program for performing this. A profile creation program used in a conversation system including a plurality of conversation devices connected to a network and sharing one of a plurality of virtual conversation spaces established on the network and capable of transmitting and receiving messages to each other is stored. A computer-readable recording medium,
Storing a predetermined keyword described in a predetermined format and a predetermined category in association with each other in the category storage means;
Storing the user, the predetermined keyword spoken by the user, the category, and the speech time of the keyword in association with each other in the user information storage unit;
A message sent to the conversation space is acquired from the conversation system, and it is determined whether or not a keyword described in a predetermined format is included in the acquired message, and a keyword described in a predetermined format is included. A category corresponding to the keyword described in the predetermined format is determined on the basis of the category storage means, further obtaining a speech time of a message including the keyword from the conversation system, and describing in the predetermined format Writing a keyword speaker, a keyword, the determined category, and the speech time in the user information storage means;
Create a conversation history in the conversation space, and at least a predetermined number n of messages (n is a positive integer) uttered before the keyword utterance or a predetermined number m (m is a positive integer) uttered after the keyword utterance Among the messages, for a message that is said within a predetermined time interval from the keyword's statement , the speaker of the message is acquired from the conversation history, and the user information is associated with the keyword as a related speaker. Further writing to the storage means;
The computer-readable recording medium which memorize | stored the profile creation program for performing this.

Images