JP2008225879A - User taste information processing method and device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To effectively utilize taste information without requiring any user's complicated management of the taste information and also without storing the taste information on a network side. <P>SOLUTION: A user terminal 10 connected to a network 50 includes: a Web page analyzing part 120 for analyzing a Web page browsed by a user; a user taste learning part 121 for automatically learning the taste of the user from a result acquired by the Web page analyzing part 120; a profile card storage part 13 for storing a profile card showing the taste of the user; and a profile card recording part 122 for generating a profile card based on the learning result of the user taste learning part 121, and stores it in the profile card storage part 13, and for updating the profile card stored in the profile card storage part 13. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a method and apparatus for processing preference information used when providing a service to a user (referred to as a user) such as a personal computer (PC) in accordance with the preference of the user. User preference information processing method that automatically analyzes and learns web (web) pages viewed by users and generates preference information, and allows users to easily register and change their preference information And an apparatus.

  With the rapid development of the Internet in recent years, the number of web pages established on the Internet has been increasing. Users of the Internet must find a web page that they want to view from a huge number of web pages. Therefore, a device for efficiently searching for a target web page has been used. Has been done. Typical ideas include (a) organizing web pages by directory (tree structure), (b) organizing web pages by keywords, and (c) organizing web pages by tagging. .

  These improvements have significantly improved the search efficiency of web pages, but many of the web page candidates presented by search engines are often uninteresting for searchers. So-called “searcher's preference” such as interest has not been considered. In recent years, there has been a return of search results based on information similar to the “searcher's preference” called “user information”, but the user has to input and specify preference information for that purpose. In addition to registration work, it is very cumbersome to maintain them regularly. In the following, problems relating to handling of preference information in web search and web services currently provided on the Internet will be specifically described.

  For example, when performing a web search, in a conventional web search, for example, as disclosed in JP-A-2006-107200 (Patent Document 1), a “popular page” or a “page most closely related to a search keyword” The search results such as "" are displayed at the top, but the page does not always match the user's preference. In order to find a web page that matches his / her preference, the user had to repeat trial and error, such as adding a search keyword, sometimes changing or deleting it.

  In recent years, various services have been provided through the web as web services, beyond the framework of simply presenting documents and images. In general web services, services such as “popular” and “popular” have been developed, and sometimes web service providers have provided services that they wanted to provide. However, web services that are deployed or provided in this way do not necessarily match individual user preferences. Japanese Patent Application Laid-Open No. 2006-252160 (Patent Document 2) and Japanese Patent Application Laid-Open No. 2004-070504 (Patent Document 3) disclose web services that register user preference information in a server and use the registered preference information. However, the freshness of the preference information is a very critical problem, and there is a problem that appropriate preference information is not always used. If the server side provides a function for updating preference information and sometimes prompts the user to periodically update the preference information, the user does not always perform the update operation. Even for the user, the continuous maintenance work of the preference information once registered is very troublesome, and this difference in consciousness between the service providing side and the user side causes a drop in service quality and a reduction in user satisfaction. In some cases, the user may not be aware of changes in the user's own preferences.

  Furthermore, there is a problem that the registered preference information can be used only within the registered web service. Therefore, when a user wants to use another web service, the user needs to register information of the same content many times for the other web service. Also, if you use a lot of web services and there are a lot of preference information registration destinations, when there is a change in the registration details, you will have to change each registration destination. Will be spent.

  In JP 2002-108918 A (Patent Document 4), preference information of a user who uses a portable information terminal can be learned according to a time zone and a place where the user performs an action. A preference learning device is disclosed in which preference information can be extracted and managed.

  Japanese Patent Laying-Open No. 2002-334274 (Patent Document 5) discloses a profile management system that centrally manages personal information (preference information) of a user. In this system, a service providing server that provides a service to a user and a profile management server that stores a user profile are interconnected via a network, and the service providing server accesses the profile management server. If the user's profile information (personal information) is acquired and the contents of the profile information are to be changed, the user accesses the profile management server via the network and requests the change. In this profile management system, it is necessary to provide a profile management server in the network separately from the user's terminal and service provision server. Since this profile management server handles the personal information of many users, strict management is required. Required.

  Japanese Patent Application Laid-Open No. 2002-342366 (Patent Document 6) describes a feature of a user's preference on a user terminal so that information desired by the user can be easily selected from information on a plurality of web pages. It is disclosed that a user profile is stored, and the user profile is automatically updated from a document vector representing characteristics of a web page viewed by the user and a browsing time of the web page by the user. The user terminal is provided with a home page selection unit that selects an information unit to be presented to the user based on the user profile.

  JP 2003-066710 A (Patent Document 7) creates and stores a user profile by analyzing a search keyword or a search sentence input by a user or a web page browsed in the past in a server computer, Furthermore, it is disclosed that when a user accesses a server computer via a network, information on the user's interests, preferences, etc. is collected to update the user profile.

  Japanese Patent Laying-Open No. 2003-122785 (Patent Document 8) discloses a user attribute registration information automatic updating apparatus that automatically updates user distribution attribute registration information using an access history to distribution information.

Japanese Patent Laying-Open No. 2003-157271 (Patent Document 9) discloses a text mining device having versatility that can be an analysis target including a web page.
JP 2006-107200 A JP 2006-252160 A JP 2004070504 A JP 2002-108918 A JP 2002-334274 A JP 2002-342366 A JP 2003-067410 A JP 2003-122785 A JP 2003-157271 A

  As described above, improvement of search accuracy and provision of a web service based on user preference are strong demands, and a web search based on preference information and an immediate solution to problems of the web service are desired.

  An object of the present invention is to provide a user preference information processing method capable of effectively using preference information without requiring complicated management of preference information by a user and without accumulating preference information on the network side. To provide an apparatus.

  A user preference information processing method according to the present invention is a user preference information processing method for processing user preference information in a user terminal connected to a network, wherein a web page viewed by a user via a network is displayed on the user terminal as a web page Profiles that indicate the user's preferences based on the results of the analysis by the analysis means, the user's preference learning means that automatically learns the user's preferences from the analysis results at the user terminal, and the learning results at the user terminal. Generating and / or updating the card by the profile card recording means.

  A user preference information processing apparatus according to the present invention is a user preference information processing apparatus that processes user preference information in a user terminal connected to a network, and that analyzes a web page browsed by a user via a network. User preference learning means for automatically learning user preferences from results obtained by web page analysis means, profile card storage means for storing profile cards indicating user preferences, and learning by user preference learning means A profile card recording means for generating a profile card based on the result of the profile and storing it in the profile card storage means and / or updating the profile card stored in the profile card storage means, and a web page Analysis method, user preference study Means, none of the profile card storage means and profile card recording means provided in a user terminal.

  That is, in the user preference information processing method of the present invention, assuming that a PC (personal computer) or the like is used as a user terminal, a web page viewed by a user (that is, a user) of the user terminal is displayed on the user terminal using a text mining technique or the like. It is used to learn what preferences the user has by automatically analyzing them, and to store the learning results in a certain common card format (called a profile card). is there. Users can use a profile card to use the Internet, which eliminates the need for complicated registration and modification of user preference information, and provides customized web services for users. Will be able to enjoy. Profile cards are automatically updated each time a user performs a web search, so that user preferences that change over time are reflected in near real time. Further, since the user can manually maintain the profile card, the freshness of the profile card can be maintained by using it together with automatic updating by a computer.

  The features of the present invention when compared with the so-called “user information” in the past are that (1) the profile card handles dynamic information of “user preference”, (2) the profile card is the user's own Managed by computer (user terminal), (3) Profile card has a certain common structure and can be used universally, (4) Profile card is automatically updated , Etc. Further, (5) the user may selectively use the profile card. By automatically updating the profile card, user preferences that change over time are reflected in the profile card in near real time, maintaining the freshness of the profile card contents and maintaining the user's profile card maintenance. It is possible to save time and effort.

  According to the present invention, in the user's terminal, the user's preference information is stored in the form of a profile card, the profile card is learned according to the user's web usage history, etc., and the profile card is based on the user's will By sending the URL to the web service providing system, etc., the user can enjoy the web service customized for the user and maintain the freshness of the profile card as the preference information. It is now possible to use preference information tailored to the user's preferences at that time, and the profile card structure is standardized, so the same profile card can be used when receiving multiple web services. There is an effect that it becomes possible.

  Next, a preferred embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing a configuration for executing a user preference information processing method according to an embodiment of the present invention.

  A user terminal 10, a web content server 20, and a web service providing system 30 are connected to a network 50 such as the Internet (or an intranet), and these can communicate with each other. The user terminal 10 is configured as a general information processing terminal such as a personal computer, and has a function of browsing (browsing) a web page. In particular, the user terminal 10 learns and stores user preference information as a profile card. It has the function to do. Specifically, such a user terminal includes a web browser 11, a user preference information learning / use unit 12 that executes processing related to learning and use of a profile card, and a profile card and storage unit 13 in which the profile card is stored. And.

  The web browser 11 includes a browsing processing unit 110 and a profile card transmission unit 111. The browsing processing unit 110 provides a web page browsing function provided in general web browser application software, and is used for browsing web pages on the Internet. The profile card transmission unit 111 also implements a data transmission function provided in a general web browser. In the present embodiment, the profile card transmission unit 111 particularly has a function of transmitting the profile card to the network 50 side.

  The user preference information learning / use unit 12 includes a web page analysis unit 120, a user preference learning unit 121, and a profile card recording unit 122. The web page analysis unit 120 has a function of analyzing the contents of the browse page when the user browses the web page using the web browser 11 and extracting a keyword included therein. is there. The user preference learning unit 121 has a function of performing learning by determining user preferences by aggregating keywords extracted by using the web page analysis unit 120 and repeating this. The profile card recording unit 122 has a function of storing the user preference information learned using the user preference learning unit 121 in the profile card storage unit 13 in the user terminal 10 as a profile card.

  In this embodiment, the profile card stored in the profile card storage unit 13 is automatically updated by the user preference information learning / use unit 12 every time a user browses a web page. The profile card is managed in the user terminal 10 and has a certain common structure (for example, a format defined in vCard). As will be described later, the user can selectively use the profile card and can manually maintain the contents of the profile card.

  The web content server 20 is a general web server that manages web pages, and includes a web page return unit 200. The web page return part 200 implement | achieves the function generally provided in the web server of returning a web page according to the request | requirement from a user.

  The web service providing system 30 is a server that provides some web service via a network 50 such as the Internet, and includes a web service providing unit 300 and a profile card interpreting unit 301. The web service providing unit 300 realizes a function provided in a general web service providing system that provides a web service in response to a request from a user. On the other hand, the profile card interpretation unit 301 has a function of interpreting the above-described profile card sent from the user terminal 10. The web service providing system 30 dynamically selects the type and contents of the service to be returned to the user based on the interpretation result by the profile card interpreting unit 301 and provides the service using the web service providing unit 300. I do.

  In this embodiment, the user preference information processing apparatus based on this invention is mounted in the user terminal which is a personal computer as the user preference information learning and utilization part 12 and the profile card storage part 13, for example. In this case, the user preference information learning / use unit 12 is realized as a computer program, and the profile card storage unit 13 is realized by setting a storage area in a storage device provided in the computer. The web browser 11 is also realized as a computer program.

  Hereinafter, the user preference information learning / use unit 12 and the profile card will be described in more detail.

  It is assumed that a computer program for realizing the user preference information learning / use unit 12 is installed in the user terminal 10 and the user preference information learning / use unit 12 starts operating in the user terminal 10. Since there is no user preference learning result at the start of operation of the user preference information learning / use unit 12, it is not known what preference the user has. In the present embodiment, the user can manually register the profile card. However, the feature of the present embodiment is that the user automatically learns the user's preference by repeating the web search. For this reason, it is assumed here that the user has made several web searches.

  When the contents of a web page browsed by the user are analyzed (for example, the text mining method disclosed in Japanese Patent Laid-Open No. 2003-157271 (Patent Document 9) is used), keywords that the user is interested in are included. Or a lot of similar information. Therefore, the user preference information learning / use unit 12 stores the keyword most frequently included in the web page browsed by the user as the user preference “candidate”. While the number of web page visits is small, it is thought that noise, that is, keywords that are not very interesting to the user, are included, but by repeating the analysis, the noise becomes relatively small and keywords that are considered to be more user-preferred Can be extracted. This is a name for averaging (continuous / addition averaging) or the like, and is a noise removal technique that is already commonly used. In the present invention, it is assumed that this technique or a technique similar thereto is used.

  When the number of web searches of a user is a certain threshold, for example, when 200 web pages are browsed, what is treated as a “candidate” is determined and stored as a profile card. This is a basic flow in which the user preference information learning / use unit 12 automatically learns user preferences and stores them in the form of profile cards. Profile that reflects the user's actual preferences by improving the accuracy of the learning results by continuously repeating this process for all web searches performed while using the user preference information learning / use unit 12 You can get a card.

  As described above, the user preference information learning / use unit 12 provides (1) a web page analysis function, (2) a user preference learning function, and (3) a profile card recording function.

  As mentioned above, users can freely maintain (register, change, add, delete) their profile cards. As a result, it is possible to improve and guarantee the accuracy of the profile card with the user's hand by supplementing functions that cannot be handled by the computer's learning ability manually, and by performing operations such as correcting incorrect learning results. To do. In addition, the profile card has a certain common structure, so that it can be easily used.

  In this embodiment, the user can selectively use the profile card. Here, “selective” has two meanings. That is, firstly, the meaning of “select profile contents” that the user can extract and use only a part of his / her profile card, and secondly, in the use of the web service, This means “select profile use” that allows the user to select whether or not to use a profile card. That is, the user can use any item on the profile card at any time.

  Hereinafter, the operation of the present embodiment will be described with reference to FIGS. FIG. 2 shows a process from analyzing and learning a user's preference based on a web page browsed by the user and storing it in the profile card, and FIG. 3 shows how the profile card is handled.

Learning user preferences:
First, user preference learning will be described.

  In user preference learning, in the user preference information learning / use unit 12, first, the web page analysis unit 120 acquires a web page browsed by the user. The web page analysis unit 120 analyzes the content of the web page using a technique such as text mining. Next, the user preference learning part 121 acquires the analysis result in the web page analysis part 120, and extracts a user preference from it. Here, by extracting user preferences every time a web page is browsed, the user preference learning unit 121 learns more accurate user preference information. When learning of user preference information is performed in this way, the profile card recording unit 122 acquires the learning result, creates a profile card based on the acquired learning result, and the profile card storage unit of the user terminal 10 13 to save.

Profile card handling:
Next, handling of the profile card will be described.

  As described above, in the present embodiment, the profile card is automatically created and updated, and the contents can be arbitrarily registered, updated, and deleted by the user. When a user intends to use a web service, the profile card can be transmitted to the web service providing system 30 using the profile card transmission unit 111 of the web browser 11 based on the user's intention. As a result, the user can enjoy a web service according to the user's preference. The web service here includes a service by a search engine.

  Next, the effect by this embodiment is demonstrated.

(1) Enjoying web services customized for you:
The user can selectively use the profile card. Therefore, when the user transmits his / her profile card to the web service providing system 30, the web service providing system 30 processes it and provides a service customized to suit individual user preferences. be able to. Since the user can control the volume of the profile card to be sent to the web service providing system 30, the range of values and tolerances are not exceeded for both users who are sensitive to the use of personal information and those who want to make effective use of personal information. You can get customized web services with maximum efficiency.

  In this way, the user can enjoy the web service customized for himself / herself only by transmitting the profile card, and the labor of trial and error found in the prior art can be saved.

(2) Maintaining the freshness of the profile card:
In this embodiment, the profile card is automatically updated. Therefore, the user's preferences that change over time are reflected in the profile card almost in real time, and the user can always use the latest preference information without being aware of the freshness. Resolving issues related to “freshness of preference information”, which was a critical issue for web services that use preference information, enables users to enjoy better services, and also provides services that match user preferences for web service providers. By doing so, it is possible to obtain effects such as reduction of opportunity loss and improvement of user satisfaction. That is, by maintaining the freshness of the profile card, the user can always enjoy a web service that suits his / her preference, and the difference between the service content and the preference, which was a problem in the prior art, is reduced.

(3) Common use of profile cards:
The profile card is managed by the user terminal, that is, the user's own computer. Moreover, the structure is a certain common one. Therefore, when the user tries to receive the web service, the user can receive the customized service immediately by transmitting the profile card, and no complicated registration procedure is required. In addition, since the profile card is updated entirely in the user's local computer, even if the profile card changes, the procedure for sending the profile card when using the web service remains the same. There is no need to take any additional action. In addition, when receiving a plurality of web services, since the structure of the profile card is made common, the user does not change the procedure such as sending the profile card. Thus, the user is free from complicated registration work and change work.

  It goes without saying that the present invention can be used in web services including services by search engines. Furthermore, the present invention can be applied in various fields by utilizing the profile card for purposes other than providing preference information to the web service providing system. For example, application to fields such as user community formation by profile card exchange and market trend survey by profile card aggregation is also possible.

It is a block diagram which shows the structure for performing the user preference information processing method of one Embodiment of this invention. It is a figure explaining a process until it analyzes and learns a user's preference based on the web page which the user browsed, and it preserve | saves to a profile card. It is a figure explaining the handling of a profile card.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 User terminal 11 Web browser 12 User preference information learning and utilization part 13 Profile card storage part 20 Web content server 30 Web service provision system 50 Network 110 Browsing process part 111 Profile card transmission part 120 Web page analysis part 121 User preference learning part 122 Profile card recording unit 200 Web page return unit 300 Web service providing unit 301 Profile card interpretation unit

Claims (6)

A user preference information processing method for processing user preference information in a user terminal connected to a network,
In the user terminal, the web page analyzing means analyzes the web page viewed by the user via the network;
In the user terminal, the user preference learning means automatically learns the user preference from the result of the analysis; and
In the user terminal, a profile card recording unit generates and / or updates a profile card indicating the user's preference based on the learning result;
A user preference information processing method.   The user preference information processing method according to claim 1, further comprising: allowing the user to maintain the profile card by himself / herself. A user preference information processing apparatus for processing user preference information in a user terminal connected to a network,
Web page analysis means for analyzing a web page viewed by the user via the network;
User preference learning means for automatically learning the user's preference from the results obtained by the web page analysis means;
Profile card storage means for storing a profile card indicating the user's preferences;
Based on the learning result of the user preference learning means, the profile card is generated and stored in the profile card storage means, and / or the profile card stored in the profile card storage means is updated. Profile card recording means,
With
The web page analyzing unit, the user preference learning unit, the profile card storage unit and the profile card recording unit are all provided in the user terminal.   4. The user preference information processing according to claim 3, further comprising: a transmission unit that is provided in the user terminal and transmits the profile card in the profile card storage unit to the network based on an instruction from the user. apparatus.   The user preference information processing apparatus according to claim 3 or 4, wherein the user can maintain the profile card in the profile card storage unit by himself / herself. Computers that make up user terminals connected to the network
Web page analysis means for a web page browsed by the user via the network;
User preference learning means for automatically learning the user's preference from the results obtained by the web page analysis means;
Profile card storage means for storing a profile card indicating the user's preferences;
Based on the learning result in the user preference learning means, the profile card is generated and stored in the profile card storage means, and / or the profile card stored in the profile card storage means is updated. , Profile card recording means,
Program to function as.

Images