JP4418135B2 - Group forming system, group forming method, and group forming apparatus - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
  The present invention relates to, for example, a group formation system and a group formation method for forming a group (hereinafter sometimes referred to as a community) in which a plurality of users on the Internet exchange information.Law andThe present invention relates to a loop forming apparatus and the like.
[0002]
[Prior art]
Currently, there are many so-called communities on the Internet, such as bulletin boards and mailing lists, where general users exchange information. Users can use the search service through a browser, enter keywords that they are interested in, perform searches, and browse information on Web pages of search results, and join communities such as bulletin boards and mailing lists on those pages. To do. We also get information about such communities from friends and acquaintances and participate in the communities.
In Japanese Patent Laid-Open No. 2000-76307, the user's terminal monitors the user's activities on the network, automatically creates the user's profile information, and when the user searches for information, the search result On the other hand, a technique for performing filtering using profile information is disclosed. In this technique, automation using profile information is performed when selecting information.
[0003]
[Problems to be solved by the invention]
However, conventionally, the formation and operation of such a community has not been automated and is performed manually.
[0004]
Specifically, in order to manage a community, it is necessary to perform the following actions.
[0005]
1. Community generation
2. Member collection
3. Community management
In order to manage the community, we will monitor the discussions within the community, delete inappropriate posts, and take steps to activate if the community falls into an inactive state. There is a need. These actions are currently performed manually, but these are very time-consuming actions.
[0006]
On the other hand, users try to find communities that they are interested in by searching. However, depending on the search technology, it is not always possible to find a community that is suitable for them. If the community felt that it was not suitable for them, the same search work had to be performed again, which was a problem.
[0007]
  The present invention is a group formation system that takes into account the above-described problems of the prior art, for example, so that the formation of a community on the Internet is less time-consuming than in the past.WhichThe purpose is to provide.
[0008]
[Means for Solving the Problems]
  1st this invention is the keyword extracted from the browsing history of the web by a browserMultiple typesAttributes and their attributesThe frequency of occurrence of keywords and the ratio of the frequency of occurrence to the total frequency of occurrence of all keywordsBuilt on the Internet, the preference information of each user with preference information having attribute values is classified as a concept, Each user belongsIn a given group,
  If a new user wants to join,New user preference information andUsing the user preference information,New user information terminals andA server connected to the information terminal of the user via the Internet is a group forming system for reconstructing a group including the new user,
  The server
  Among the preference information of users belonging to the predetermined plurality of groups,Keyword appearance frequency shows a stable appearance frequency for a certain period of timeA first database storing first preference information;
  Among the preference information of users belonging to the predetermined plurality of groups,The appearance frequency of the keyword shows a large degree of change compared to the change degree of the appearance frequency of the keyword of the first preference information in a certain period.A second database storing second preference information;
  A classification concept database storing the data of the predetermined groups constructed on the Internet, wherein the preference information of each user is classified as a concept;
  By the COBWEB methodThe first preference information of the new user and the preference information of each user of the classification concept databaseCU values representing similarity in the group and dissimilarity between the groups are obtained from the attribute and the attribute value, and the group is selected so that the CU value is maximized.Reconstructing the predetermined plurality of groups;
  By the COBWEB methodThe second preference information of the new user and the preference information of each user of the classification concept databaseCU values representing similarity in the group and dissimilarity between the groups are obtained from the attribute and the attribute value, and the group is selected so that the CU value is maximized.A classification unit for reconstructing the predetermined plurality of groups;
  And a classification information transmitting unit that transmits the reconstructed result to a user information terminal.
[0009]
  The second aspect of the present invention is a keyword extracted from a web browsing history by a browser.Multiple typesAttributes and their attributesThe frequency of occurrence of keywords and the ratio of the frequency of occurrence to the total frequency of occurrence of all keywordsBuilt on the Internet, the preference information of each user with preference information having attribute values is classified as a concept, Each user belongsIn a given group,
  If a new user wants to join,New user preference information andUsing the user preference information,New user information terminals andA server connected to the user's information terminal via the Internet is a group formation method for reconstructing a group including the new user,
  Among the preference information of users belonging to the predetermined group, the server isKeyword appearance frequency shows a stable appearance frequency for a certain period of timeStoring first preference information in a first database in the server;
  Among the preference information of users belonging to the predetermined group, the server isThe appearance frequency of the keyword shows a large degree of change compared to the change degree of the appearance frequency of the keyword of the first preference information in a certain period.Storing second preference information in a second database in the server;
  The server classification partBy the COBWEB methodIn the server, the first preference information of the new user and the preference information of each user are classified as a concept, and data of the predetermined plurality of groups constructed on the Internet are stored.MinutesThe preference information of each user in the similar concept databaseCU values representing similarity in the group and dissimilarity between the groups are obtained from the attribute and the attribute value, and the group is selected so that the CU value is maximized.Reconstructing the predetermined plurality of groups;
  By the COBWEB methodThe second preference information of the new user and the preference information of each user of the classification concept databaseCU values representing similarity in the group and dissimilarity between the groups are obtained from the attribute and the attribute value, and the group is selected so that the CU value is maximized.Reconstructing the predetermined plurality of groups;
  A classification information transmission unit of the server, the step of transmitting the reconstructed result to a user information terminal;
  A group forming method characterized by comprising:
[0010]
  The third aspect of the present invention is a keyword extracted from a web browsing history by a browser.Multiple typesAttributes and their attributesThe frequency of occurrence of keywords and the ratio of the frequency of occurrence to the total frequency of occurrence of all keywordsBuilt on the Internet, the preference information of each user with preference information having attribute values is classified as a concept, Each user belongsIn a given group,
  If a new user wants to join,New user preference information andReconstructing the group including the new user using the user preference information;New user information terminals andA group forming apparatus connected to the information terminal of the user via the Internet,
  Among the preference information of users belonging to the predetermined plurality of groups,Keyword appearance frequency shows a stable appearance frequency for a certain period of timeA first database storing first preference information;
  Among the preference information of users belonging to the predetermined plurality of groups,The appearance frequency of the keyword shows a large degree of change compared to the change degree of the appearance frequency of the keyword of the first preference information in a certain period.A second database storing second preference information;
  A classification concept database storing the data of the predetermined groups constructed on the Internet, wherein the preference information of each user is classified as a concept;
Information from
  By the COBWEB methodThe first preference information of the new user and the preference information of each user of the classification concept databaseCU values representing similarity in the group and dissimilarity between the groups are obtained from the attribute and the attribute value, and the group is selected so that the CU value is maximized.Reconstructing the predetermined plurality of groups;
  By the COBWEB methodThe second preference information of the new user and the preference information of each user of the classification concept databaseCU values representing similarity in the group and dissimilarity between the groups are obtained from the attribute and the attribute value, and the group is selected so that the CU value is maximized.A classification unit for reconstructing the predetermined plurality of groups;
  A classification information transmission unit that transmits the reconstructed result to a user information terminal;
  A group forming apparatus.
[0011]
  The fourth aspect of the present invention is a keyword extracted from a web browsing history by a browser.Multiple typesAttributes and their attributesThe frequency of occurrence of keywords and the ratio of the frequency of occurrence to the total frequency of occurrence of all keywordsBuilt on the Internet, the preference information of each user with preference information having attribute values is classified as a concept, Each user belongsIn a given group,
  If a new user wants to join,New user preference information andUsing the user preference information,New user information terminals andA server connected to the user's information terminal via the Internet is a group formation method for reconstructing a group including the new user,
  Among the preference information of users belonging to the predetermined group, the server isKeyword appearance frequency shows a stable appearance frequency for a certain period of timeA first database in the server for storing first preference information;
  Among the preference information of users belonging to the predetermined group, the server isThe appearance frequency of the keyword shows a large degree of change compared to the change degree of the appearance frequency of the keyword of the first preference information in a certain period.A second database in the server for storing second preference information;
  Using the classification concept database in the server, the server storing the data of the predetermined plurality of groups constructed on the Internet, wherein the preference information of each user is classified as a concept,
  The server classification partBy the COBWEB methodThe first preference information of the new user and the preference information of each user of the classification concept databaseCU values representing similarity in the group and dissimilarity between the groups are obtained from the attribute and the attribute value, and the group is selected so that the CU value is maximized.Reconstructing the predetermined plurality of groups;
  By the COBWEB methodThe second preference information of the new user and the preference information of each user of the classification concept databaseCU values representing similarity in the group and dissimilarity between the groups are obtained from the attribute and the attribute value, and the group is selected so that the CU value is maximized.Reconstructing the predetermined plurality of groups;
  A classification information transmission unit of the server, the step of transmitting the reconstructed result to a user information terminal;
  A group forming method characterized by comprising:
[0043]
DETAILED DESCRIPTION OF THE INVENTION
  Less than,The present invention and related to the present inventionEmbodiments of the present invention will be described with reference to the drawings.
[0044]
FIG. 1 is a diagram showing the overall system configuration.
[0045]
In FIG. 1, reference numeral 101 denotes a network such as the Internet.
[0046]
102 and 104 are users who are participating in or intending to participate in the community, and 103 and 105 collect preference information necessary to determine which community is appropriate when the user joins the community. , An information terminal for generating.
[0047]
A server 106 manages the community, receives user preference information transmitted from the users 102 and 104, and performs clustering of users based on the accumulated preference information of the plurality of users and the received preference information. And see the created class as a community. Thereafter, information indicating which community the user belongs to is transmitted to the user.
[0048]
Reference numeral 107 denotes a service provider, which provides the user with product information, service information, and the like related to the content discussed in the community. An advertiser 108 provides advertisement information to the community in which the user participates.
[0049]
(Embodiment 1)
In the first embodiment, user preference information is automatically extracted by an information terminal, data having a digital signature and an authentication function is transmitted to the server, and the server authenticates the received data. A procedure for extracting user preference information data from received data will be described.
[0050]
FIG. 2 is a diagram in which the user's preference is extracted from information that the user is browsing the browser and data is transmitted.
[0051]
In FIG. 2, 201 is a browser such as Internet Explorer, and 202 is a language processing unit that performs language processing such as morphological analysis on language information viewed by the browser 201.
[0052]
A feature extraction unit 203 extracts data representing user preferences from the data processed by the language processing unit 202.
[0053]
Reference numeral 204 denotes a personal information database that stores information related to individual users such as the user's gender, age, and hobbies.
[0054]
Reference numeral 205 denotes a preference history storage database that stores preference information when the user already holds preference information.
[0055]
Reference numeral 206 denotes user preference information output from the feature extraction unit 203 regarding the content of the browser being browsed by the user, personal information output from the personal information database 204, and preference history output from the preference history storage database 205. It is a preference creation unit that creates new preference information of a user from information.
[0056]
Reference numeral 207 denotes a transmission data creation database that stores the data format when the preference information is transmitted as transmission data.
[0057]
Reference numeral 208 denotes a key database that stores a key used for assigning a digital signature.
[0058]
Reference numeral 209 denotes a signature unit that signs data using the key of the key database 208.
[0059]
210 performs tagging on the preference information created by the preference creation unit in accordance with the format stored in the transmission data creation database 207, and passes the tagged information to the signature unit 209. Is a transmission data creation unit that creates a transmission data by adding a signature to the tagged data.
[0060]
A data transmission unit 211 transmits the data created by the transmission data creation unit 210.
[0061]
Next, the operation when the language data shown in FIG. 4 is input through the browser will be described with reference to the flowchart of FIG.
[0062]
In step 301, corresponding to the operation of the language processing unit 202, the language data 401 is subjected to morphological analysis to obtain a morphological analysis result as indicated by reference numeral 501 in FIG.
[0063]
Step 302 corresponds to the operation of the feature extraction unit 203, extracts only nouns from the data 501 subjected to morphological analysis, and creates data composed of the vocabulary as shown by 601 in FIG. 6 and its appearance frequency.
[0064]
Steps 303 to 309 correspond to the operation of the preference creation unit 206, select one word at a time from the noun data 601, refer to the data 701 stored in the preference history storage database 205, and select the noun. It is determined whether it is already stored in the database 205. As shown in the preference history data 701 in FIG. 7, the preference history storage database 205 stores the history information of the current personal preference, the word as an attribute, the appearance frequency viewed by the user, and the appearance frequency. A ratio (weight) of the total appearance frequency of all words is stored as an attribute value.
[0065]
Looking at the word “professional baseball” in the extracted noun data 601 (see FIG. 6), “professional baseball” is stored in the preference history data 701 (see FIG. 7), so the appearance frequency addition in step 306 is executed. Then, the appearance frequency is added once, and the appearance frequency becomes “11 times”.
[0066]
In step 309, it is determined whether there is a noun that is not selected in the noun data 601. If there is a noun, the next noun “Japan” and “series” are sequentially selected in steps 303 to 305, and the preference history data 701 is stored. With reference to it, it is determined whether it is already stored.
[0067]
Here, a case where the last noun “free agent” in the noun data 601 is selected will be considered.
[0068]
In step 304, it is determined with reference to the preference history data 701 that “free agent” has not appeared so far. In step 307 and step 308, the word “free agent” and the appearance frequency “once” are newly registered as attributes in the preference history data 701. In step 309, it is determined that there is no unselected word in the noun data 601.
[0069]
In step 310 and step 311, the weight of each word as an attribute is calculated as an attribute value from all the appearance frequencies obtained by adding the noun data 601 to the preference history data 701, and new preference history data shown in 801 of FIG. 8 is created. To do.
[0070]
In step 312, personal information data as shown in FIG. 9 stored in the personal information database 204 is assigned to the created new preference history data 801, and the preference information data 1001 shown in FIG. Create as preference information.
[0071]
The transmission data creation unit 210 receives the preference information 1001 created by the preference creation unit 206 according to the flowchart described above. The transmission data creation unit 210 creates transmission data in a tagged format as indicated by 1101 in FIG. For example, XML can be used as a tagging rule.
[0072]
In the figure, reference numeral 1102 denotes “Profile” information of transmission data. Details of the “Profile” information, as described in 1201 of FIG. 12, “transmission destination IP address” that represents the transmission destination of transmission data, “transmission source IP address” that represents the transmission source that transmitted the transmission data, It consists of the “name” of the user and the “signature” applied by the signature unit 209 to the transmission data.
[0073]
In FIG. 11, 1103 describes a method that can be executed on the receiving side as “Action” information for transmission data. Details of the “Action” information are composed of a method for extracting preference information included in the transmission data and a method for extracting personal information included in the transmission data, as described in 1202 of FIG. .
[0074]
In FIG. 11, the contents portion 1104 surrounded by the <Contents> tag describes the preference information 1001 created by the preference creation unit 206.
[0075]
The tag names of Agent, Profile, Action, and Content may use other names and symbols as long as they can be uniquely identified on the transmission side and the reception side.
[0076]
Tagged data and tagging data 1301 as shown in FIG. 13 are created from the transmission data creation database 207 and the transmission data creation unit 210.
[0077]
For the operation in the signature unit 209, “Pair” “Private Key” and “Public Key” are created in advance for data authentication, and “Private Key” is stored in the key database 208. On the other hand, the “public key” is stored in the server that receives the transmission data.
[0078]
Here, these “pair” keys may be created at the user's information terminal, and the “public key” may be sent to the server, or the key may be created on the server side, and the “private key” It may be downloaded or stored in the user's key database 208 using another storage medium. In order to improve security, a public key and a private key may be newly created every predetermined period.
[0079]
A procedure for adding a “signature” to the tagging data 1301 and creating the transmission data 1405 will be described with reference to FIG.
[0080]
In FIG. 14, reference numeral 1401 denotes a key creation unit, which creates a “pair” “private key” and “public key” for data authentication.
[0081]
1402 is a private key stored in the key database 208, and 1403 is a public key stored in the server.
[0082]
The signature unit 209 receives the tagging data 1301 from the transmission data creation unit 210, extracts the “private key” 1402 from the key database 208, and creates a “signature”.
[0083]
The “signature” created by the signature unit 209 is returned to the transmission data creation unit 210, and the transmission data creation unit 210 adds “signature” 1404 to the tagging data 1301, creates the transmission data 1405, and creates data The transmission data 1405 shown in FIG.
[0084]
Next, a procedure in which the server 106 (see FIG. 1) receives transmission data 1405 transmitted from the user information terminal, performs authentication, and extracts user preference information from the transmission data 1405 will be described.
[0085]
FIG. 16 is a diagram modeling this procedure.
[0086]
In FIG. 16, reference numeral 1601 denotes a data receiving unit that receives transmission data 1405 transmitted from the user's information terminal.
[0087]
A method database 1603 stores methods that can be executed by the server for received data.
[0088]
Reference numeral 1604 denotes a registrant information management database that manages user information registered in the server.
[0089]
Reference numeral 1602 denotes an authentication rule storage database that stores rules for authenticating the data 1405 received by the data receiving unit 1601.
[0090]
A data authentication unit 1605 refers to the registrant information management database 1604 and the authentication rule storage database 1602 with respect to the received data 1405 and performs authentication using the methods stored in the method database 1603.
[0091]
A preference information extraction unit 1606 extracts preference information by referring to a method for extracting user preference information from data authenticated by the data authentication unit 1605 among methods stored in the method database 1603. is there.
[0092]
The authentication procedure will be described with reference to the flowchart shown in FIG. 17 and FIG.
[0093]
Step 1701 corresponds to the operation of the data receiving unit 1601 and receives data transmitted by the user.
[0094]
Steps 1702 to 1704 correspond to the operation of the data authentication unit 1605. A method of the method database 1603 used in the data authentication unit 1605 will be described with reference to FIG. Reference numeral 1801 denotes a signature authentication method for authenticating a signature described in received data. Reference numeral 1802 denotes a format authentication method for authenticating the format of received data. Reference numeral 1803 denotes a user information authentication method for authenticating user information described in received data.
[0095]
First, in step 1702, whether or not the signature of the received data 1405 is correct by using the public key 1403 (created by the key creation unit 1401 in FIG. 14) stored in the authentication rule storage database 1602 by the signature authentication method 1801. Authenticate. If the signature matches, in step 1703, the format authentication method 1802 authenticates the received data 1405 using the data 1901 describing the format that the received data should have, which is also stored in the authentication rule storage database 1602. Do. In the format data 1901, the format to be included in the received data is that the outermost frame is surrounded by “<Agent>” and “</ Agent>” tags, and “<Profile>”, “</ Profile>”, “ This indicates that data surrounded by “<Action>” and “</ Action>”, “<Contents>” and “</ Contents>” are described. If the format is correct, in step 1704, the user information authentication method 1803 determines whether the user data 1902 managed in the registrant information management database 1604 matches the data in the received data 1405. FIG. 20 shows details of the user data 1902.
[0096]
Of the data managed in the user data 1902 in FIG. 20, it is determined whether or not the user name and password match the “Profile” information in the received data 1405.
[0097]
In step 1705, if the received data 1405 is not authenticated in any of steps 1702 to 1704, error processing is performed.
[0098]
Step 1706 corresponds to the operation of the preference information extraction unit 1606 in FIG. 16, and uses the preference information extraction method 1804 that is a method stored in the method database 1603 for the data authenticated in steps 1702 to 1704. Then, user preference information is extracted.
[0099]
In this embodiment, it is assumed that keywords are extracted from web browsing history by a browser as data for creating user preference information. However, the present invention is not limited to this. It is also conceivable to use data such as purchase history and trends of goods at the store.
[0100]
By using the purchase history, for example, people who recently purchased baby items such as diapers and powdered milk may have various concerns, anxiety, and consultation regarding child-rearing and childbirth, so such people will participate It can be used to create a community.
[0101]
(Embodiment 2)
In the second embodiment, a method of creating a community when a new participant appears when a community has already been established on the Internet will be described.
[0102]
That is, a community (hereinafter sometimes referred to as a category) constructed based on the preference information 1001 of a user who wants to newly participate and the preference information of a plurality of users who have already participated, received by the server 106. ) On how to create a community with new users by clustering from the information about.
[0103]
Of the server configuration shown in FIG. 16, components used in this embodiment will be described.
[0104]
Reference numeral 1607 denotes a classification concept database that stores concept information as a result of the current clustering.
[0105]
Reference numeral 1608 denotes a preference accumulation database that stores preference information as shown in FIG. 10 regarding all members belonging to the community.
[0106]
Reference numeral 1609 denotes a classification unit that performs new clustering with reference to the preference accumulation database 1608 and the classification concept database 1607 including the preference information 1001 that is the output of the preference information extraction unit 1606 in the first embodiment.
[0107]
Reference numeral 1610 denotes a classification information transmission unit that refers to the registration information management database 1604 and notifies a change to a user whose category has changed as a result of new classification by the classification unit 1609.
[0108]
Next, the classification concept database 1607 will be described. Details are shown in FIG.
[0109]
In the figure, an ellipse indicates a node indicating a category branch (corresponding to 2101).
[0110]
A square indicates a category (corresponding to 2102), and the preference information of each user belonging to the category indicates that the variance is smaller than a certain value.
[0111]
In the figure, members belonging to one category “Class: 3” 2103 are “User1”, “User7”, “User13”, “User25”, “User28”, “User30”, “User39”, “User42”. It shows that there is.
[0112]
The class to which each member belongs is stored in order from the upper node. Taking “User1” as an example, it is stored that it belongs to “Class: 3”, which is subordinate to “Node: 2” and subordinate to “Node: 5”.
[0113]
Next, an operation when the classification unit 1609 receives new preference information as an input will be described.
[0114]
The case where the concept forming method “COBWEB” that automatically forms a concept from a plurality of cases is used as the classification unit 1609 will be described. The details of “COBWEB” are entrusted to the document “Fisher. Conceptual clustering, Learning from Examples, and Inference, Proceeding of the Fourth International Workshop on Machine Learning, pp. 38-49, 1987”.
[0115]
In “COBWEB”, a value called a category utility (hereinafter referred to as CU value) is used as an index for classification. The CU value is a value that represents the similarity of the data in the category and the dissimilarity of the data between the categories, and the operation of the classification unit 1609 proceeds while selecting the conceptual configuration that maximizes the CU value.
[0116]
The classification unit 1609 performs the following four operations on the new preference information and the current clustering result (corresponding to FIG. 21).
[0117]
As for the operation, the operation for determining the optimum category for the hierarchy one level lower than “Node: 1” in FIG. 21 will be described.
[0118]
1. Add to current category
As shown in the example of 2201 in FIG. 22A, CU values are calculated when new preference information is added to existing categories “Node: 2”, “Node: 3”, and “Node: 4”, respectively. To do.
[0119]
2. Create new category
As indicated by 2202 in FIG. 22A, a CU value is calculated when a category is newly created in the same row as “Node: 2”, “Node: 3”, and “Node: 4”.
[0120]
3. Category integration
As shown in an example at 2301 in FIG. 23A, CU values are calculated for all cases where two categories (Nodes 2 and 3 in FIG. 21) are integrated into a new category (new Node in FIG. 23A).
[0121]
4). Category separation
As shown in an example at 2302 in FIG. 23B, CU values are calculated for all cases where one category (Node 3 in FIG. 21) is separated and two categories (Class 5 and 6 in FIG. 23B) are created.
[0122]
The above four operations are performed, the classification method having the largest CU value is adopted, and the same operation is performed for the lower category.
[0123]
The classification information transmission unit 1610 receives the output of the classification unit 1609, and the user newly added to the category and the user to which the category to which the user belongs as a result of the classification by the classification unit 1609 have been changed. The new category information is notified by referring to the IP address and Mail Address of the user data 1902 shown in FIG. 20 in the registrant information management database 1604.
[0124]
In the present embodiment, the case where a new user participates in the community has been described. However, the present invention is not limited to this. For example, the following is also possible. That is, when the preference information of a user who has already participated is changed, the changed preference information is handled in the same manner as the preference information of the new user in this embodiment, so that the user whose preference has changed newly belongs to the community. It is also possible to make it.
[0125]
(Embodiment 3)
In the third embodiment, a procedure for allowing a user belonging to a certain community to belong to one or more other communities will be described.
[0126]
FIG. 24 is a block diagram for explaining the operation in the present embodiment.
[0127]
A preference history database 2401 stores preference history of each user.
[0128]
Reference numeral 2402 denotes a preference history reflection data creation unit that creates preference information that reflects history with reference to history information in the preference history database 2401.
[0129]
The purposes of the preference history database 2401 and the preference history reflection data creation unit 2402 are described below.
[0130]
The preference data stored in the preference accumulation database 1608 for a certain individual can be regarded as data representing the tendency of interest in the long-term span of the individual. On the other hand, by storing the transition of the long-term preference data in the preference history database 2401, preference information representing a tendency of interest in a short-term span can be created.
[0131]
That is, when the tendency of interest is expressed by the attribute and its frequency (weight), long-term preference data (corresponding to the first preference information of the present invention) indicating the frequency of stable appearance in a certain period, and the above-mentioned certain period Short-term preference data (corresponding to the second preference information of the present invention) showing a greater degree of change than the change in the frequency of appearance of long-term preference data can be created.
[0132]
Then, among the preference information data stored in the preference storage database 1608, the community clustered by the long-term preference data and the community clustered by the short-term preference data are different from each other.
[0133]
This allows a single user to belong to multiple communities, and any community that has a long-term interest of the user and a short-term interest that the user will be interested in, such as “Recent Trends” Can also participate.
[0134]
FIG. 25 shows data representing a history stored in the preference history database 2401.
[0135]
As shown in the figure, the attribute value for each attribute whose attribute value is digitized by a weight represents a change in the attribute value for each individual stored in the preference accumulation database 1608.
[0136]
The preference history reflection data creation unit 2402 creates data that prospered the interest of recent users as shown in FIG. 26 from the data in the preference history database 2501 shown in FIG. In this data, the attribute value of the attribute whose plus / minus value is plus is set high, and the attribute value of the minus attribute is set low.
[0137]
Thereby, preference information data reflecting the characteristics of recent preferences can be created.
[0138]
The classification unit 1609 receives the output data 2601 of the preference history reflection data creation unit 2402, the output of the classification concept database 1607, and the preference accumulation database 1608, and performs the operation shown in the second embodiment, and performs the operation shown in FIG. Determine the category to which the data belongs. Thus, the user can use the community clustered by the long-term span preference data stored in the preference accumulation database 1608 and the short-term span preference data created by the preference history database 2401 and the preference history reflection data creation unit 2402. It becomes possible to belong to a clustered community and two communities.
[0139]
The classification information transmission unit 1610 notifies the user of a new category in the same manner as described in the second embodiment.
[0140]
The preference data for clustering users into a plurality of communities is not limited to data based on the history of preference, and the preference information may be newly created by selecting attributes from preference information according to some selection criteria. Good.
Moreover, the preference data representing the short-term interest of the user as shown in the present embodiment may be created by a server that manages the community, or may be created by an information terminal that the user usually uses.
[0141]
(Embodiment 4)
In the fourth embodiment, an operation for reorganizing a community (category reorganization) for an inactive community where information exchange is not actively performed will be described.
[0142]
FIG. 27 is a block diagram for explaining the operation in the present embodiment.
[0143]
In the figure, reference numeral 2702 denotes a posted information storage database that stores the sender, date, and information content of information posted in the community.
[0144]
Reference numeral 2701 denotes an activity determination unit that refers to data stored in the posted information storage database 2702 and determines the activation level of the community.
[0145]
The activity determination unit 2701 extracts a community that does not satisfy the activity determination criteria from the posted information storage database 2702.
[0146]
Examples of the activity determination criterion include “the number of postings per day”, “the number of fixed posters”, and “the number of community members” as shown in FIG.
[0147]
The classification unit 1609 receives information on an inactive community (category) from the activity determination unit 2701 and performs integration with an activated community or integration and reorganization of inactive communities.
[0148]
In the present embodiment, a method for automatically detecting and reorganizing a community that has not been activated has been described. However, by using an activity determination criterion as shown in FIG. It is possible to present in a form that can be easily identified according to the degree of activation.
[0149]
For example, in FIG. 49, information such as which community is currently being discussed most actively by changing the size of the graphic or character that points to each community's activity level on the portal site of the community. Can be conveyed visually. Another possible method is to change the color of the figure, character, or wallpaper instead of the size of the figure or character.
[0150]
Further, in the portal site as shown in FIG. 49, the positional relationship of graphics, characters, etc. pointing to each may be displayed in correlation with the degree of similarity between communities.
[0151]
(Embodiment 5)
In the fifth embodiment, a case where a company provides a service regarding contents exchanged in a certain community will be described.
[0152]
First, FIG. 29 shows a configuration diagram used in this embodiment.
In the figure, reference numeral 2901 denotes a posted information monitoring unit that monitors and extracts information posted by users in the community.
[0153]
Reference numeral 2902 denotes a provided information storage database in which information that the service providing company 107 wants to provide to users, such as product and service information, is registered together with keywords that are paired with the information.
[0154]
Reference numeral 2903 denotes a keyword collating unit that collates a keyword included in the information extracted by the posted information monitoring unit 2901 with a keyword registered in the provided information storage database 2902 to determine whether or not they match.
[0155]
2904 searches the provided information storage database 2902 for information that is paired with the keyword determined to be matched by the keyword matching unit 2903, and posts the paired information to the user who posted the keyword or the keyword. It is a provision information display part displayed with respect to the user who participates in the community including a user.
[0156]
Next, a processing procedure in the present embodiment will be described.
[0157]
FIG. 30 shows data posted to the posted information storage unit 2702. FIG. 31 shows an example of information registered in the provided information storage database 2902. In this example, a URL that the user wants to refer to is described as information that the service provider wants to provide for the keyword posted by the user.
[0158]
  Assume that information shown in 3001 is posted in the posted information storage database 2702. The posted information monitoring unit 2901 monitors the posted information storage database 2702 and extracts 3001 information. The keyword collation unit 2903 collates whether or not the information stored in the provided information storage database 2902 is included in the information extracted by the posted information monitoring unit 2901. In this embodiment, “Digital camera", The provided information display unit 2904 searches the provided information storage database for the keyword"Digital camera"Http://www.marumarudenki.co.jp/digi-came", which is information corresponding to " To display. The user accesses the displayed URL,Digital cameraCan be obtained.
[0159]
In this embodiment, URL information such as products and services is described as information provided to the user by the service provider. However, the information provided is not a URL, but a direct description such as a description or price of the product or service. 32, or explanatory information related to the keyword posted by the user, as indicated by 3201 in FIG.
[0160]
(Embodiment 6)
In the sixth embodiment, a case will be described in which an advertiser displays an advertisement related to the content exchanged in the community.
[0161]
First, FIG. 33 shows a configuration diagram used in this embodiment.
[0162]
In the figure, a posted information monitoring unit 2901 and a posted information storage database 2702 have the same functions as the elements shown in the fifth embodiment.
[0163]
The keyword storage database 3301 is a means for storing keywords included in information posted to the community together with the number of posts.
[0164]
The keyword adding unit 3302 extracts a keyword from the information extracted by the posted information monitoring unit 2901. If the keyword is stored in the keyword storage database 3301, the number of posts is added. If the keyword is not stored, the keyword is added. This is a means for newly storing the number of postings.
[0165]
Reference numeral 3303 denotes an advertiser database that stores an advertiser who wants to display an advertisement to a community together with a display condition and displayed advertisement content. As a condition for displaying, it is considered that the number of keywords posted to the community reaches a certain number or more.
[0166]
Reference numeral 3304 denotes an advertisement display determination unit that searches for a keyword stored in the keyword storage database 3301 and determines whether the number of postings satisfies an advertiser condition in the advertiser database 3303.
[0167]
Reference numeral 3305 denotes an advertisement display unit that displays an advertiser's advertisement determined to be matched by the advertisement display determination unit 3304 in the community.
[0168]
Next, a processing procedure in the present embodiment will be described.
[0169]
Reference numeral 3401 in FIG. 34 denotes information stored in the keyword storage database 3301. The keyword and the number of posts are stored in pairs. 3402 is a diagram showing a result of adding a keyword and the number of posts by the operation of the keyword adding unit 3302. Reference numeral 3501 in FIG. 35 denotes information stored in the advertiser database 3303, in which a company name, advertisement contents of the company, a keyword that is a condition for displaying the advertisement of the company, and the number of posts are stored in pairs. ing.
[0170]
As in the fifth embodiment, it is assumed that information indicated by 3001 in FIG. 30 is posted to the posted information storage database 2702 by the user. The posted information monitoring unit 2901 monitors the posted information storage database 2702 and extracts 3001 information.
[0171]
  The keyword adding unit 3302 receives the output of the posted information monitoring unit 2901 and uses “JPEG”, “image”, “Digital camera”Is extracted, and the number of posts of each keyword is added to the data indicated by 3401 in FIG. 34 to create data indicated by 3402.
[0172]
  The advertisement display determination unit 3304 determines a company whose data in the keyword storage database 3301 satisfies an advertisement display condition among companies stored in the advertiser database 3303. In this embodiment, the condition of “Company A” out of 3501 data, “JPEG is 5 times”, “Digital cameraIt is determined that “10 times” is satisfied.
[0173]
The advertisement display unit 3305 receives the output of the advertisement display determination unit 3304 and displays the advertisement of “Company A” to the community.
[0174]
In this embodiment, the condition for displaying the advertisement is the number of postings of the keyword. However, the appearance ratio of a specific keyword for all keywords may be used, or an advertisement may be displayed according to the desire of a company analyzing information posted to the community. It may be displayed.
(Embodiment 7)
In the seventh embodiment, with respect to keywords and contents included in information posted to the community, it is necessary to refer to user profiles (preference information, personal information, etc.) belonging to the community and understand the contents and meaning A method for presenting explanation information to a likely user will be described.
[0175]
First, FIG. 36 shows a configuration diagram in the present embodiment.
[0176]
In the figure, 3602 is a concept storage database that stores the conceptual structure of words such as words and compound words.
3603 is a profile management unit that stores personal profiles such as preference information and personal information of users belonging to the community.
3601 extracts keywords, such as words and compound words, from the posted information to the community monitored by the posted information monitoring unit 2901, and refers to the concept storage database 3602 and the profile management unit 3603 to determine the keywords. It is a display information determination unit that identifies a set of users and keywords that need to be explained with regard to the contents.
Reference numeral 3604 denotes an explanation information storage database that stores keywords such as words and compound words and explanation items in pairs.
[0177]
FIG. 38 is a diagram showing an example of the concept storage database 3602. In this example, the concept is expressed in a tree structure, and is composed of categories (words surrounded by squares in the figure) and attributes (words included in the lowermost category in the figure). Categories such as “hobby”, “learning”, and “practical” are in the same hierarchy, and subcategories of “hobby” include “sports”, “music”, etc. There is “baseball”. The “sports” and “music” categories are both subordinate to the “hobby” category, so they are closely related to each other. However, the “IT-related products” category is a subordinate concept to the “practical” category, so it is subordinate to the “hobby” category. I think that it is not closely related to the concept of “sports” and “music” categories.
[0178]
Next, the flow of processing will be described using a specific example.
[0179]
The posted information monitoring unit 2901 detects that the information shown in FIG. 37 has been posted to the community. The display information determination unit 3601 needs to provide detailed explanation information regarding which keyword to which user with reference to the profile of the community member user stored in the profile management unit 3603 and the concept storage database 3602. Determine whether.
[0180]
Here, a method for determining whether or not it is necessary to present keyword explanation information to a user having a certain profile will be described.
[0181]
Assume that a user has a profile as shown in FIG. From the attribute of the profile, the user is interested in the “sports” category of the conceptual structure of FIG. 38 and is considered to have knowledge about this category.
[0182]
  When words and compound words are extracted from the posted information in FIG.Digital camera"Is the word. "Digital camera”In the conceptual structure of FIG. 38, it can be seen that it belongs to the“ IT related products ”category. At this time, the display information determination unit 3601 displays “Digital cameraSince the “IT related products” category to which “” belongs and the “sports” category in which the user is interested are not closely related, it is determined that the user has little knowledge about the attributes belonging to the “IT related products” category. Therefore, for the user having the profile shown in FIG. 10, when the information shown in FIG. 37 is posted or when the user browses the information,Digital cameraThe information 3901 shown in FIG. 39A is presented to the user as the explanation information. Information 3901 in FIG. 39A is stored in the explanation information storage database 3604.
[0183]
The above determination is made for all users belonging to the community, and information 3901 shown in FIG. 39A is presented to a user who seems to have little knowledge about the posted keyword as judged from the profile. Can do.
[0184]
The explanatory information as shown in FIG. 39 may be displayed in the browser of the community being browsed by the user (see FIG. 39B), or may be displayed in a separate window. May be.
(Embodiment 8)
In the eighth embodiment, a method will be described in which community posting information is monitored, and communities that have similar topics are integrated into one community.
[0185]
FIG. 40 is a diagram showing a configuration in the present embodiment.
[0186]
In the figure, reference numerals 4001 to 4003 denote posting information storage databases for storing topics posted to community 1, community 2, and community N, respectively.
[0187]
A post information extraction unit 4004 refers to the post information storage databases 4001 to 4003 and extracts information posted to each community.
[0188]
Reference numeral 4005 denotes a conceptual structure database representing the conceptual structure of vocabulary.
[0189]
Reference numeral 4007 denotes a community similarity storage unit that stores the similarity of information posted between communities.
[0190]
Reference numeral 4006 refers to information posted to each community obtained by the posted information extraction unit 4004 and the concept similarity database 4005, and adds the similarity between communities stored in the community similarity storage unit 4007. It is a degree addition part.
[0191]
Reference numeral 4008 denotes a similar community determination unit that determines similar communities based on the similarity of the community similarity storage unit 4007.
[0192]
FIG. 41 shows an example of the conceptual structure stored in the conceptual structure database 4005.
[0193]
The conceptual structure is expressed in a hierarchical structure between vocabularies as shown in the figure. Under the concept of “music”, it means that the concepts of “Western music”, “Japanese music”, “female vocals, and“ jazz ”exist. To do. In this example, we think that the vocabulary enclosed in a square represents a community. For example, “music group B” and “music group C” in “Japanese music” characterize the “Japanese music community”. Think of it as a keyword.
[0194]
FIG. 42 shows an example of information stored in the community similarity storage unit 4007.
[0195]
The numerical value in the figure indicates that, for example, a keyword characterizing “female vocal community” shown in the conceptual structure of FIG. 41 has appeared 18 times in the information posted so far to “Japanese music community”. In other words, in the “Japanese music community”, keywords related to “female vocal community” have been posted 18 times in the past. The larger this number, the deeper the relationship between the two communities is, and the higher the similarity is. .
[0196]
Next, the flow of processing in the present embodiment is shown. Note that FIG. 41 is adopted as the conceptual structure database 4005 and FIG. 42 is adopted as the community similarity storage unit 4007.
[0197]
When the information shown in FIG. 43 is posted to the “Japanese music community”, the posted information extraction unit 4004 extracts the information.
[0198]
The community similarity addition unit 4006 refers to the conceptual structure database 4006 in FIG. 41 and adds similarities between communities. The post information to “Japanese music community” shown in FIG. 43 includes the keywords “female vocal A” and “female vocal B” related to “female vocal community” once. The similarity to “female vocal community” is added by “2”, and the data in the community similarity storage unit 4007 in FIG. 42 is changed from “18” to “20” as shown in FIG.
[0199]
The similar community determination unit 4008 monitors the community similarity storage unit 4007 and determines whether there is a community that satisfies the standard for community integration. As a criterion for the determination, for example, “the numerical value of the similarity between communities stored in the community similarity storage unit 4007 is“ 20 ”or more” can be considered.
[0200]
In FIG. 44, the similarity between the “Japanese music community” and the “female vocal community” satisfies the above-described criteria, so the similar community determination unit 4008 determines to integrate the “Japanese music community” and the “female vocal community”.
[0201]
It is desirable to notify members belonging to the community that the community has been integrated.
(Embodiment 9)
In the ninth embodiment, a method will be described in which information posted in a community is monitored and a new community is created independently from an existing community according to the posted content.
[0202]
FIG. 45 is a diagram showing a configuration in the present embodiment.
[0203]
In the figure, reference numeral 4501 denotes a posted information storage database.
Reference numeral 4503 denotes a keyword storage unit that stores, for each member in the community, a keyword included in information posted to the community and the number of postings.
[0204]
Reference numeral 4502 denotes a keyword addition unit that extracts a keyword from post information to the community extracted by the post information extraction unit 4004 and adds the number of keyword postings to the keyword storage unit 4503.
Reference numeral 4504 denotes an independent community determination unit that monitors the keyword storage unit 4503 and determines whether an independent community is newly generated according to the number of keyword postings.
[0205]
FIG. 46 shows a specific example of the keyword storage unit 4503.
[0206]
As shown in the figure, the keyword storage unit 4503 is managed for each member belonging to “music community” in the conceptual structure database 4005 of FIG. 41, and each member is “Western music” which is a subordinate concept of “music community”. "," Japanese music ", and the number of times of posting the keywords (shown in FIG. 41) characterizing" female vocal "are stored. For example, the member “Jiro” of the “music community” indicates that sentences including keywords such as “musician group A” and “artist A” belonging to the concept “Western music” have been posted four times.
[0207]
Next, the flow of processing in the present embodiment will be described.
[0208]
As a precondition, it is assumed that a “music community” currently exists and members belonging to the community are shown in the keyword storage unit 4503 in FIG.
[0209]
When the text shown in FIG. 47 is posted in the posted information storage database 4501, the posted information extracting unit 4004 extracts the information and transmits it to the keyword adding unit 4502.
[0210]
The keyword adding unit 4502 refers to the conceptual structure database 4005 in FIG. 41, and each member adds keywords related to the concepts of “Western music”, “Japanese music”, “female vocal”, and “jazz” below the “music community”. Determine if you are posting those keywords.
[0211]
In this example, keywords such as “Artist B” of member “Midori”, “Artist D” of “Jun”, “Musician group B” of “Sam” are determined as keywords corresponding to the above concept, and FIG. The number of keywords in the keyword storage unit 4503 is incremented by 1 for each member, resulting in the number of keywords as shown in FIG.
[0212]
When a keyword is added to the keyword storage unit 4503, the independent community determination unit 4504 refers to the keyword storage unit 4503 to determine whether there is a community that is newly independent from the “music community” that is the existing community. For example, the following can be considered as a criterion.
[0213]
“If a member who belongs to a certain community A and has posted a keyword related to the subordinate concept B of the concept A representing the community A more than a predetermined number of times, if there are more than a predetermined number of members, a new Create community B. "
In the above criteria, the “predetermined number of times” is 20 times, the “predetermined number of people” is 4, and the keyword storage unit 4503 updated as shown in FIG. 48 shows that “Japanese music” satisfies the above criteria. . Accordingly, the independent community determination unit 4504 determines to generate a new community “Japanese music community” with “Sam”, “Midori”, “Maki”, and “Jun” who are members satisfying the above criteria.
[0214]
The member may belong only to the new “Japanese music community” or may belong to the “music community” and two communities.
[0215]
In the present embodiment, as a community to be independent, “Western music” and “Japanese music” for “music” and subordinate concepts for the existing community are listed as candidates as shown in the example of FIG. 41. Not only concepts that have a “parent-child” relationship but also different concepts may be cited as candidates for independent communities.
[0216]
In the above embodiment, the case where the information terminal (corresponding to the terminal device of the present invention) collects and generates user preference information and the like has been mainly described. However, the present invention is not limited to this, and for example, connected to the Internet. A configuration performed by a server (community server 106 in FIG. 1) may be used.
[0217]
In the above embodiment, the group forming apparatus of the present invention has been described centering on the case where it is realized by a single community server 106. However, the present invention is not limited to this. You may do it.
[0218]
The present invention is a program for causing a computer to execute the functions of all or part of the above-described group forming system and group forming apparatus of the present invention (or apparatus, element, circuit, unit, etc.). The program operates in cooperation with the computer.
[0219]
The present invention is also a program for causing a computer to execute all or some of the steps (or processes, operations, actions, etc.) of the above-described group forming method of the present invention. It is a program that operates.
[0220]
Further, the present invention is a medium carrying a program for causing a computer to execute all or part of the functions of all or part of the above-described group forming system and group forming apparatus of the present invention. It is a medium that can be read and the read program executes the function in cooperation with the computer.
[0221]
In addition, the present invention is a medium carrying a program for causing a computer to execute all or part of the steps of the above-described group forming method of the present invention. The program is a medium for executing the operation in cooperation with the computer.
[0222]
Further, some means (or devices, elements, circuits, parts, etc.) of the present invention and some steps (or processes, operations, actions, etc.) of the present invention are a plurality of these means or steps. It is intended to mean some means or steps within, or to mean some functions or operations within one means or step.
[0223]
The present invention also includes a computer-readable recording medium that records the program of the present invention.
[0224]
Further, one usage form of the program of the present invention may be an aspect in which the program is recorded on a computer-readable recording medium and operates in cooperation with the computer.
[0225]
Further, one usage form of the program of the present invention may be an aspect in which the program is transmitted through a transmission medium, read by a computer, and operated in cooperation with the computer.
[0226]
The recording medium includes a ROM and the like, and the transmission medium includes a transmission medium such as the Internet, light, radio waves, sound waves, and the like.
[0227]
The computer of the present invention described above is not limited to pure hardware such as a CPU, but may include firmware, an OS, and peripheral devices.
[0228]
As described above, the configuration of the present invention may be realized in software or hardware.
[0229]
As is clear from the above description, using the present invention, a user who wants to participate in a community can automatically search from his / her preference information without having to search for a community suitable for him / her by using a search engine or the like. Can find a suitable community. Further, when one community is not suitable for the user, another community can be automatically found. Furthermore, when the community is not activated, reorganization can be automatically performed without bothering people.
[0230]
On the other hand, a company that wants to provide services and products to the community can provide users with appropriate advertising information at the appropriate time without having to constantly monitor the opinions and information exchanged in the community. It is possible to appropriately display explanatory texts and images for users who are difficult to understand in the posted information.
[0231]
In the above-described embodiment, the preference information may include purchase information for an article.
[0232]
【The invention's effect】
  As is clear from the above description, the present invention is, for example, a community on the Internet.Formation,It has the advantage of being easier to perform than in the prior art.
In addition, it is possible to easily provide explanations for terms that are not relevant to users belonging to the community.
[Brief description of the drawings]
FIG. 1 is a system configuration diagram according to the present embodiment.
FIG. 2 is a configuration diagram of the information terminal according to the present embodiment.
FIG. 3 is a flowchart of creating preference information in the information terminal.
FIG. 4 shows language data
FIG. 5 is a diagram showing a morphological analysis result of language data.
FIG. 6 is a diagram showing extracted nouns
FIG. 7 is a view showing preference history data.
FIG. 8 is a diagram showing updated preference history data
FIG. 9 is a diagram showing personal information data
FIG. 10 is a diagram showing preference information data
FIG. 11 is a diagram showing a transmission data format
FIG. 12 is a diagram showing data for creating transmission data
FIG. 13 is a view showing preference information data tagged.
FIG. 14 shows a signature creation procedure.
FIG. 15 is a diagram showing transmission data of an information terminal and reception data of a community server
FIG. 16 is a configuration diagram of a community server.
FIG. 17 is a flowchart of an authentication procedure.
FIG. 18 is a diagram showing data in a method database
FIG. 19 shows the authentication procedure.
FIG. 20 is a diagram showing registered user data.
FIG. 21 is a conceptual diagram of clustered categories.
FIG. 22A is a diagram showing a clustering operation (adding new preference information in each existing category). FIG. 22B is a clustering operation (creating a new category in the same row as an existing category). Figure
FIG. 23A shows a clustering operation (category integration). FIG. 23B shows a clustering operation (category separation).
FIG. 24 is a configuration diagram of a community server in the third embodiment.
FIG. 25 is a diagram showing data stored in a preference history database
FIG. 26 is a diagram showing preference information created by reflecting the preference history.
27 is a configuration diagram of a community server in Embodiment 4. FIG.
FIG. 28 is a diagram showing an activity determination criterion
FIG. 29 is a configuration diagram of a community server in the fifth embodiment.
FIG. 30 is a diagram showing information posted from the user to the community;
FIG. 31 is a diagram showing an example of data stored in a provided information storage database
FIG. 32 is a diagram showing an example of data stored in the provided information storage database
FIG. 33 is a configuration diagram of a community server in the sixth embodiment.
FIG. 34 is a diagram showing data stored in a keyword storage database.
FIG. 35 is a diagram showing data stored in an advertiser database.
FIG. 36 is a block diagram of a community server in the seventh embodiment.
FIG. 37 is a diagram illustrating an example of information posted to a community
FIG. 38 is a diagram showing tree structure data expressing the concept of vocabulary.
FIGS. 39A and 39B are diagrams showing vocabulary explanation information on the community site.
FIG. 40 is a configuration diagram of a community server in the eighth embodiment.
FIG. 41 is a diagram showing a conceptual tree structure of vocabulary and keywords characterizing the vocabulary
FIG. 42 is a diagram illustrating an example of data stored in a community similarity storage unit
FIG. 43 is a diagram showing an example of information posted to a community site
FIG. 44 is a diagram illustrating an example of data stored in a community similarity storage unit
FIG. 45 is a configuration diagram of a community server in the ninth embodiment.
FIG. 46 is a diagram illustrating an example of data stored in a keyword storage unit
FIG. 47 is a diagram showing an example of information posted to a community site
FIG. 48 is a diagram illustrating an example of data stored in a keyword storage unit
FIG. 49 shows an example of a community portal site
[Explanation of symbols]
101 Internet
102 User 1
103 Information terminal 1
104 users 2
105 Information terminal 2
106 Community server
107 Service provider
108 Advertiser
201 browser
202 Language processor
203 Feature extraction unit
204 Personal information database
205 Preference history storage database
206 Preference Creation Department
207 Database for creating transmission data
208 key database
209 Signature part
210 Transmission data generator
211 Data transmitter
301 Morphological analysis step
302 Noun extraction step
303 noun word selection step
304 History reference step
305 Pre-existing noun confirmation step
306 Appearance frequency addition step
307 Vocabulary registration step
308 Appearance frequency registration step
309 Other noun presence confirmation step
310 Weight calculation step
311 History update step
312 Step for giving personal information
401 Language data
501 Morphological analysis results
601 Noun extraction data
701 Preference history data
801 Updated preference history data
901 Personal information data
1001 Preference information
1101 Transmission data format
1102 Profile description part
1103 Action description part
1104 Contents description part
1201 Profile details
1202 Action details
1301 Tagged preference data
1401 Key creation unit
1402 Private key
1403 Public key
1404 Signature
1405 Transmission data
1601 Data receiver
1602 Authentication rule storage database
1603 Method Database
1604 Registrant Information Management Database
1605 Data authentication unit
1606 preference information extraction unit
1607 Classification concept database
1608 Preference accumulation database
1609 Classification part
1610 Classification information transmission unit
1701 Data reception step
1702 Signature conformance confirmation step
1703 Format conformance confirmation step
1704 Registrant Identification Step
1705 Error handling step
1706 Taste information acquisition step
1801 Signature authentication method
1802 Format authentication method
1803 User information authentication method
1804 Preference information extraction method
1901 Format of received data
1902 User data
2101 Example of a node indicating a branch of clustering
2102 Example of clustering category
2103 Clustering categories
2201 Operation to add to current category
2202 Operation to create a new category
2301 Operations for integrating categories
2302 Operation to separate categories
2401 Preference history database
2402 Preference history reflection data creation unit
2501 Data representing changes in user preferences
2601 Preference information data created reflecting history
2701 Activity determination unit
2702 Posting information storage database
2801 Activity Criteria
2901 Post Information Monitoring Unit
2902 Provided information storage database
2903 Keyword matching part
2904 Provided information display section
3001 Post data from users
3101 Data example of provided information storage database
3201 Data example of provided information storage database
3301 Keyword storage database data
3302 Keyword addition unit
3303 Advertiser database data
3304 Advertisement display determination unit
3305 Advertisement display unit
3401 Keyword storage database data before addition
3402 Keyword storage database data after addition
3501 Advertiser database data
3601 Display information determination unit
3602 Concept memory database
3603 Profile Management Department
3604 Explanation information storage database
4001 Posting information storage database 1
4002 Posting information storage database 2
4003 Posting information storage database N
4004 Posting information extraction unit
4005 Conceptual structure database
4006 Community similarity adder
4007 Community similarity storage
4008 Similar community judgment unit
4501 Posting information storage database
4502 Keyword adder
4503 Keyword storage
4504 Independent community judgment section

Claims (4)

Preference that has multiple types of attributes that are keywords extracted from the browsing history of the web by the browser, the frequency of appearance of the keywords of that attribute , and the attribute value that indicates the ratio of the appearance frequency to the total appearance frequency of all keywords In the plurality of predetermined groups to which each user belongs, constructed on the Internet, wherein the preference information of each user having information is classified as a concept.
When a new user wishes to participate, a new user information terminal and a server connected to the user information terminal via the Internet using the new user preference information and the user preference information, A group formation system for reconstructing a group including the new user,
The server
Of the preference information of users belonging to the plurality of predetermined groups, a first database storing first preference information in which the appearance frequency of a keyword indicates a stable appearance frequency in a certain period ;
A second preference in which, among the preference information of users belonging to the plurality of predetermined groups, a keyword appearance frequency shows a greater degree of change than a change degree of a keyword appearance frequency in the first preference information over a certain period of time A second database storing information;
A classification concept database storing the data of the predetermined groups constructed on the Internet, wherein the preference information of each user is classified as a concept;
By the COBWEB method , the similarity in the group and the dissimilarity between the groups are expressed from the first preference information of the new user and the attributes and attribute values of the preference information of each user in the classification concept database. Obtaining a CU value, selecting a group so that the CU value is maximized, and reconstructing the plurality of predetermined groups;
By the COBWEB method, the second preference information of the new user and the attribute and the attribute value of the preference information of each user of the classification concept database represent the similarity in the group and the dissimilarity between the groups. A classification unit that obtains a CU value, selects a group so as to maximize the CU value, and reconstructs the predetermined plurality of groups;
A group formation system comprising: a classification information transmission unit that transmits the reconstructed result to a user information terminal. Preference that has multiple types of attributes that are keywords extracted from the browsing history of the web by the browser, the frequency of appearance of the keywords of that attribute , and the attribute value that indicates the ratio of the appearance frequency to the total appearance frequency of all keywords In the plurality of predetermined groups to which each user belongs, constructed on the Internet, wherein the preference information of each user having information is classified as a concept.
When a new user wishes to participate, a new user information terminal and a server connected to the user information terminal via the Internet using the new user preference information and the user preference information, A group formation method for reconstructing a group including the new user,
The server stores, in a first database in the server, first preference information indicating that the appearance frequency of a keyword is stable in a certain period of user preference information belonging to the plurality of predetermined groups. And steps to
The server shows a degree of change in the preference frequency of keywords among the preference information of the users belonging to the plurality of predetermined groups as compared to a change rate of the frequency of appearance of the keywords of the first preference information in a certain period. Storing second preference information in a second database in the server;
The classification unit of the server is configured such that the first plurality of preference information of the new user and the preference information of each user are classified as concepts according to a COBWEB method . storing a group of data, determine the CU value representing the dissimilarity between the similarity and the group in the group from the attribute and the attribute value of the preference information of each user minute Ruigainen database in the server, the Selecting a group to maximize the CU value and reconstructing the predetermined plurality of groups;
By the COBWEB method, the second preference information of the new user and the attribute and the attribute value of the preference information of each user of the classification concept database represent the similarity in the group and the dissimilarity between the groups. Obtaining a CU value, selecting a group such that the CU value is maximized, and reconstructing the predetermined plurality of groups;
A classification information transmission unit of the server, the step of transmitting the reconstructed result to a user information terminal;
A group forming method comprising: Preference that has multiple types of attributes that are keywords extracted from the browsing history of the web by the browser, the frequency of appearance of the keywords of that attribute , and the attribute value that indicates the ratio of the appearance frequency to the total appearance frequency of all keywords In the plurality of predetermined groups to which each user belongs, constructed on the Internet, wherein the preference information of each user having information is classified as a concept.
When a new user wishes to participate, the new user information terminal and the user information are reconstructed by using the new user preference information and the user preference information to reconstruct the group including the new user. A group forming device connected to a terminal via the Internet,
Of the preference information of users belonging to the plurality of predetermined groups, a first database storing first preference information in which the appearance frequency of a keyword indicates a stable appearance frequency in a certain period ;
A second preference in which, among the preference information of users belonging to the plurality of predetermined groups, a keyword appearance frequency shows a greater degree of change than a change degree of a keyword appearance frequency in the first preference information over a certain period of time A second database storing information;
A classification concept database storing the data of the predetermined groups constructed on the Internet, wherein the preference information of each user is classified as a concept;
Information from
By the COBWEB method , the similarity in the group and the dissimilarity between the groups are expressed from the first preference information of the new user and the attributes and attribute values of the preference information of each user in the classification concept database. Obtaining a CU value, selecting a group so that the CU value is maximized, and reconstructing the plurality of predetermined groups;
By the COBWEB method, the second preference information of the new user and the attribute and the attribute value of the preference information of each user of the classification concept database represent the similarity in the group and the dissimilarity between the groups. A classification unit that obtains a CU value, selects a group so as to maximize the CU value, and reconstructs the predetermined plurality of groups;
A classification information transmission unit that transmits the reconstructed result to a user information terminal;
A group forming apparatus comprising: Preference that has multiple types of attributes that are keywords extracted from the browsing history of the web by the browser, the frequency of appearance of the keywords of that attribute , and the attribute value that indicates the ratio of the appearance frequency to the total appearance frequency of all keywords In the plurality of predetermined groups to which each user belongs, constructed on the Internet, wherein the preference information of each user having information is classified as a concept.
When a new user wishes to participate, a new user information terminal and a server connected to the user information terminal via the Internet using the new user preference information and the user preference information, A group formation method for reconstructing a group including the new user,
The first database in the server in which the server stores first preference information indicating the appearance frequency in which the appearance frequency of a keyword is stable in a certain period among the preference information of users belonging to the predetermined groups. When,
The server shows a degree of change in the preference frequency of keywords among the preference information of the users belonging to the plurality of predetermined groups as compared to a change rate of the frequency of appearance of the keywords of the first preference information in a certain period. A second database in the server for storing second preference information;
Using the classification concept database in the server, the server storing the data of the predetermined plurality of groups constructed on the Internet, wherein the preference information of each user is classified as a concept,
The classification unit of the server uses the COBWEB method to calculate the similarity and group within the group from the first preference information of the new user and the attributes and attribute values of the preference information of each user of the classification concept database. CU values representing dissimilarities between them are selected, a group is selected so that the CU value is maximized, and the predetermined plurality of groups are reconstructed,
By the COBWEB method, the second preference information of the new user and the attribute and the attribute value of the preference information of each user of the classification concept database represent the similarity in the group and the dissimilarity between the groups. Obtaining a CU value, selecting a group such that the CU value is maximized, and reconstructing the predetermined plurality of groups;
A classification information transmission unit of the server, the step of transmitting the reconstructed result to a user information terminal;
A group forming method comprising:

Images