JP2009193465A - Information processing apparatus, information providing system, information processing method, and program - Google Patents

Abstract

Information is provided to a larger number of users by using a network such as the Internet.
For example, a cluster processing unit 104 uses a URL of a Web page collected by a user terminal device 131 accessing a Web page prepared on the Internet 120 as action specifying information, and performs an action By classifying users (user IDs) whose records (behavior history) of a plurality of accessed URLs (behavior identification information) stored in the history storage unit 103 are similar, identifiers of contents accessed by these users A cluster is generated from (content ID). In addition, the cluster processing unit 104 associates a user (user ID) classified as similar to the generated cluster.
[Selection] Figure 1

Description

  The present invention relates to an information processing apparatus, an information providing system, an information processing method, and a program that provide a user with information estimated to be related based on user behavior such as an access history with respect to a network such as the Internet.

  In recent years, companies and consumers (individual users) that provide information on stores and products using the Internet are increasing. A typical example is an advertisement. In addition, there are an increasing number of sites that publish information written by individuals, such as blogs (weblogs) and SNSs (Social Network Service), so that more people can view their sites. Some users have devised to increase the chances of browsing by leaving a footprint, such as adding a trackback and writing comments. In order to effectively provide such information, it is important to narrow down the users who are applicable to the information to be provided, for example, to purchase behavior. As a method for narrowing down such users, there is a technique of grasping user characteristics from user preference information collected using a network and grouping similar characteristics together (see Patent Document 1). ).

  In this technology, when there are "concept vectors associated with a large number of words", "user profile created from user service history", "information that the information provider (advertisement provider) wants to send (product advertisement) ) Is expressed by a concept vector, and the user of the user profile having a concept vector close to that of the advertised product is advertised (product information is provided).

  In addition to the techniques described above, as a related technique, there is an information filtering technique that extracts user characteristics in the form of user interest or connection between products. For example, an example of online shopping is given. These use collaborative filtering technology to compare users with the same purchase history compared to product sales history, and "People who have bought these items also bought these items." By displaying (providing) such information on the terminal, it recommends products that have not been bought so far.

JP 2004-126975 A

  However, first, in the technique of Patent Document 1, it is necessary to prepare a concept vector in advance, and this concept vector is expressed as a dimension vector, and what information such as news, newspapers, and dictionaries is used. Since the accuracy and characteristics depend on whether the concept vector was created based on the above, there is a problem that it cannot be used easily. For example, in the technique of Patent Document 1, it is possible to express the relationship between words with high accuracy as the vector dimension of the concept vector increases, but there is a problem that the calculation amount becomes enormous as the vector dimension increases. is there.

  In addition, in the creation of concept vectors, characteristic words are extracted from the user's service usage history, user profiles are created from the extracted words, and information that information providers want to provide is selected as characteristic keywords. There is a need to. Since the information provider is not an engineer of this method, it is not easy to determine which keyword is effective. In addition, when a word is extracted from a user's service usage history, text analysis is performed, but it cannot be applied to a video such as a photograph or a video.

  Moreover, in the service using collaborative filtering, although it is suitable as recommended information for product sales, the target user is restricted to online shopping users. In this case, for example, a user who does not perform online shopping cannot be provided with information about the product or the store, such as advertising the product or the store.

  In addition, in the related technology described above, a user who matches (matches) information such as a user's hobbies and preferences such as advertising and specific information such as a user's hobbies and preferences, or a purchased product provides information. There is a problem that information is not provided to users who are the target and do not match.

  To summarize the above, conventionally, there are a limited number of users to whom information is provided, and there is a problem that it is not easy to provide information to more users.

  The present invention has been made in view of the above points, and an object thereof is to provide information to more users by using a network such as the Internet.

  An information processing apparatus according to the present invention includes an action history storage unit in which action identification information obtained from a state in which a terminal device is connected to a network is stored as an action history together with a user ID for identifying a user, and the action history storage A cluster processing unit that classifies user IDs with similar behavior histories stored in the unit, and generates a cluster of behavior identification information included in the behavior history of the classified user ID in association with the classified user ID; At least a collation processing unit that selects a cluster including the set action specifying information from the clusters generated by the cluster processing unit and extracts a user ID associated with the selected cluster as an information providing target.

  In the information processing apparatus, the cluster processing unit classifies user IDs with similar temporal transitions of behavior specifying information included in the behavior history. The behavior history storage unit stores behavior identification information as a behavior history together with time information in addition to the user ID, and the cluster processing unit generates a cluster from the behavior identification information whose time information is in a predetermined time zone. . In addition, the cluster processing unit generates a cluster from the action specifying information equal to or more than a predetermined number of overlaps over the action history of each classified user ID from the action specifying information included in the action history of the classified user ID.

  In the information processing apparatus, the collation processing unit excludes a user who includes at least a part of information to be provided in the action history from the information provision target.

  In the information processing apparatus, sending of an e-mail to the user terminal apparatus corresponding to the user ID extracted by the collation processing unit as information provision target with the information associated with the action specifying information used for selecting the cluster A plurality of means including a display on a Web page provided at a site accessed by the user terminal device and a display on a Web page provided at a site established by a user identified by the user ID Provide by means selected from among. Further, the information provision control unit provides information by a selected means in a preset order.

  An information providing system according to the present invention is an information providing system that is connected to an information provider terminal and a user terminal and provides information registered by the information provider terminal to the user terminal. Actions registered with information from the information provider terminal from a plurality of clusters reflecting the similarity of action histories associated with the user terminal device or the user identification information and configured from action specifying information acquired by It comprises at least means for selecting a cluster including specific information, selecting a user terminal associated with the selected cluster, and notifying means for notifying information to the selected user terminal.

  In addition, the information processing method according to the present invention includes a user ID having a similar action history including action specifying information that is configured from information obtained from a state in which the terminal device is connected to the network and indicates a state of use of the network by the user. Is generated by associating a cluster of action specifying information included in the action history of the classified user ID with the classified user ID, and setting the action specified from the clusters generated by the cluster processing unit. In this method, a cluster including information is selected, a user ID associated with the selected cluster is extracted, and the user ID extracted by the collation processing unit is set as an information providing target.

  In addition, the program according to the present invention includes a user ID that identifies a user with action specifying information that is configured from information obtained from a state in which a terminal device is connected to a network and that indicates a state of use of the network by the user. A cluster composed of action identification information included in the action history of the classified user ID by classifying user IDs similar to the action history stored in the action history storage function and the action history storage function stored as the action history A cluster generation processing function that is generated in association with a user ID that is classified, and a cluster that includes the set action specifying information is selected from the clusters generated by the cluster generation processing function, and is associated with the selected cluster. A program for realizing a verification processing function for extracting a user ID .

  As described above, according to the present invention, the behaviors included in the behavior history of the classified user ID are classified by classifying the user IDs having similar behavior histories including behavior specifying information indicating the state of use of the network by the user. Since a cluster composed of specific information is generated and information is provided to users by using this cluster, an excellent effect that information can be provided to more users can be obtained.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a configuration diagram showing a configuration example of an information providing system for realizing an information processing method according to an embodiment of the present invention. This system includes an information processing device 100, a user terminal device 131, a mobile terminal device 132, and an information provider terminal 140 that can be connected to the information processing device 100 via the Internet 120. The user terminal device 131 is connected to the Internet 120 via an Internet service provider (ISP) 121. The mobile terminal device 132 is connected to the Internet 120 via the ISP 122.

  In addition, a service providing server 150 is connected to the Internet 120, and services on the Internet such as online shopping, search, video distribution, net banking, and online payment information are available, and Web information can be browsed. It is possible. Further, the present invention is not limited to the Internet 120, and may be another wide area network.

  Here, the information processing apparatus 100 is, for example, a server provided in a portal operator or ISP that provides services on the web including various information provisions called a so-called portal site. The information processing apparatus 100 may be a server provided in a service provider that can accumulate and manage user behavior histories. The information provider terminal 140 is a terminal provided in an information provider such as an advertiser who performs an advertisement or the like. The information provider, for example, owns information (contents) such as advertisements to be provided to the user and link (access destination) information thereof.

  Hereinafter, a case will be described in which the information processing apparatus 100 is provided in a portal business operator and the information provider terminal 140 is provided in an advertiser. For example, the user 130 is a service providing server 150 operated by a portal company, for example, via the ISP 121, ISP 122, the Internet 120, etc., by a computer, a mobile phone, a net TV, etc. The service on the Internet is used, and Web information is browsed. In addition, the advertiser registers an advertisement to be provided to the user with the portal operator. Correspondingly, the portal provider provides information such as advertising to each user.

  As described above, the information processing apparatus 100 that provides information is first, for example, a user who stores information on a user who is registered (connection is permitted) in the service providing server 150 operated by the portal provider. Information storage unit 101 and an action history processing unit that generates various action histories obtained from a state in which a user (user terminal device 131, portable terminal device 132) stored in user information storage unit 101 is connected to a network 102. The action history processing unit 102 collects action specifying information obtained from a state in which the user is connected to the network together with a user ID for identifying the user, and generates an action history in which the collected action specifying information is recorded.

  The behavior specifying information includes, for example, information indicating the state of use of the network by the user. In each service provided by the service providing server 150 in which the user identified by the user ID is registered, These include the above information (URL: Uniform Resource Locator), photos, reproduced videos, purchased product information, links to these, and access times to the services provided. Further, as the action specifying information, information transmitted from the mobile terminal device 132 connected to the Internet 120 can be used. For example, if the identification information (telephone number) of the mobile terminal device 132 by the user 130 is registered, the call time of the mobile terminal device 132, the travel location (position information) sent from the mobile terminal device 132, the stay time, etc. It can be used as action specific information.

The action history processing unit 102 collects action specifying information by, for example, leaving a record of access to the service providing server 150 of the user terminal device 131. The same applies to the portable terminal device 132. In addition, when the information processing apparatus 100 is placed in a provider such as an ISP, the user terminal apparatus collects action specifying information by leaving a record when connecting to the Internet via the ISP. Can do. In the collection of such action specifying information, it is possible to record the Web information (URL) browsed by the user and the information of the site visited by the user, and this information can be handled as the action specifying information.
Also, information recorded when an IC tag (RFID tag) provided in a mobile terminal device or a card possessed by the user is connected to a communication device such as a reader installed in a place prepared in advance. Thus, for example, position information that specifies a place where the communication device is installed can be collected as action specifying information.

  Information obtained from commercial transactions on the Internet can also be used as action specifying information. For example, the behavior identification of purchase history in online shopping carried out by portal operators and information on traded products obtained by collecting information recorded when these business transactions are made by credit card payment. It can be used as information. In addition to these, from users who have obtained permission, exchange of e-mails can be accumulated, and product information or the like can be used as action specifying information from the contents of the accumulated e-mails.

  In addition, when an account aggregation service is performed in which the portal operator performs personal information management in one stop, the portal operator (information processing apparatus 100 by the user terminal device 131, for example, by the user 130). The usage history of a plurality of services such as online shopping malls and information distribution services via) is also accumulated as action specifying information. This account aggregation service is a mechanism that allows you to receive corporate services without revealing personal information to service providers by registering personal information with portal providers when receiving services from companies that provide services on the Internet. It is. Normally, it is necessary for the user to provide personal information to the service provider, but this is not necessary by performing this account aggregation. Furthermore, in account aggregation, online banking and payments by credit card payment may be affiliated with related companies, and such information can also be stored as action specifying information.

  As described above, the action history processing unit 102 collects action specifying information from various types of information obtained by the techniques described above from a state in which the action history processing unit 102 is connected to the network in various forms. More action history. For example, the action history includes action specifying information including the URL of the browsed Web page and information on the date and time of browsing (accessing). Note that the action history is not limited to only one of the above-described information forms (behavior specifying information). For example, a combination of all types of action specifying information or a combination of some action specifying information may be mixed to form an action history.

  In addition, the information processing apparatus 100 includes an action history storage unit 103 that stores an action history. For example, as described above, an action history including action specifying information and user ID collected by the action history processing unit 102 is stored in the action history storage unit 103. Note that the action history stored in the action history storage unit 103 is not limited to the action history information collected by the action history processing unit 102. For example, an action history configured from action specifying information collected by other means may be stored in the action history storage unit 103.

  Further, the information processing apparatus 100 classifies user IDs having similar behavior histories from the behavior histories stored in the behavior history storage unit 103, and thus includes behavior specifying information included in the behavior histories of the classified user IDs. A cluster processing unit 104 that generates a cluster in association with the classified user ID and a cluster storage unit 105 that stores a cluster generated by the cluster processing unit 104 are provided.

  For example, the cluster processing unit 104 uses the URL of the Web page collected when the user terminal device 131 accesses a Web page prepared on the Internet 120 as action specifying information, and uses the action history storage unit 103. By classifying users (user IDs) whose records (behavior history) of a plurality of accessed URLs (behavior identification information) stored in are similar, action identification information (for example, URLs) accessed by these users Create a cluster from Here, the cluster is generated by grouping or classifying, and the action specifying information included in the action history of each user is compared between the users, and the similar user and the action specifying information of this user are compared. Are grouped into one group. In addition, the cluster processing unit 104 associates a user (user ID) classified as similar to the generated cluster. For example, a cluster is generated with the action specifying information and the user ID based on the classification described above.

  In addition, the information processing apparatus 100 includes a provision information storage unit 106 that stores a provision information ID that identifies information to be provided in association with the behavior identification information, and a behavior identification stored in the provision information storage unit 106. The cluster storage unit 105 is searched using the information as a key, a corresponding cluster is selected, and a matching processing unit 107 that acquires a user ID associated with the selected cluster is provided. The provided information ID and the action specifying information stored in the provided information storage unit 106 are, for example, selected and set by an advertiser who wants to advertise to the user. The provided information ID may be any information that can identify information to be provided, such as a URL of a Web page on which the provided information is posted. Further, for example, the provided information ID and the action specifying information are transmitted from the information provider terminal 140 and received by the transmission / reception unit 109.

  In addition, the information processing apparatus 100 retrieves the provided information ID associated with the action specifying information used by the collation processing unit 107 for selecting the cluster from the provided information storage unit 106, and collates based on the retrieved provided information ID. The information provision control unit 108 that generates provision information for the user corresponding to the user ID extracted by the processing unit 107, and the provision information generated by the information provision control unit 108 is sent to the target user terminal device 131 and portable terminal device 132. Transmitting and receiving unit 109.

  Next, user information stored in the user information storage unit 101 will be described. The user information storage unit 101 stores, for each user, registered user information including a user's name, whereabouts, age, gender, e-mail address, authentication code, password, and the like together with a user ID for identifying the user. Yes. Here, the user information storage unit 101 only needs to store at least a user ID for identifying a user to whom information is to be provided. The authentication code may be used as the user ID.

  Next, the action history stored in the action history storage unit 103 will be described. In the action history storage unit 103, for example, as shown in Table 1 below, the access date and time, the action specifying information for identifying the accessed Web page, and the access information including the user ID are stored as the action history.

Table 1
Access date Access start time Action specific information User ID
20071001 11:00:13 cid0010 uid0001
20071001 11:01:01 cid0010 uid0003
20071001 11:01:45 cid0005 uid0001
20071001 11:05:01 cid0015 uid0002
20071001 11:08:00 cid0005 uid0003
20071001 11:08:01 cid0001 uid0002
20071001 11:09:25 cid0015 uid0003
::::

  In Table 1, for example, a user whose “user ID” is “uid0001” has an “access start time” whose “access date” is “20071001” is “11:00:13”, and “cid0010” as action specifying information is displayed. ”Is shown. “Cid0001”, “cid0005”, “cid0010”,... Are URLs and product IDs of accessed Web pages, for example, and are simply shown for the sake of simplicity.

  Next, clusters stored in the cluster storage unit 105 will be described. The cluster storage unit 105 stores, for example, clusters having action specifying information as shown in Table 2 below. Here, a case where each cluster includes a corresponding user ID is shown. In Table 2, for example, the cluster 1 includes “cid0005”, “cid0010”, and “cid0015” as the action specifying information, and also includes user IDs “uid0001” and “uid0003”. ing.

Table 2
Cluster 1: cid0005, cid0010, cid0015, ... | uid0001, uid0003
Cluster 2: cid0002, cid0003, cid0005, ... | uid0012
Cluster 3: cid0009, | uid0023
Cluster 4: cid0001, cid0015 | uid0002, uid0019
:::

  Next, information to be provided and associated action specifying information stored in the provided information storage unit 106 will be described. The provided information storage unit 106 stores, for example, action specifying information as shown in Table 3 below and a provided information ID for identifying provided information. This is information specified in advance by the advertiser, for example. Note that these do not need to be prepared in advance, and may be prepared, for example, at the stage of advertisement. In Table 3, for example, “inf0001” is associated as the provided information ID with “cid0005” as the action specifying information. Note that “inf0001”, “inf0001”, and “inf0003” are, for example, URLs of Web pages on which advertisements are posted, and are simply shown for easy explanation.

Table 3
Action specific information: Provided information ID
cid0005: inf0001
cid0003 or cid0015: inf0002
cid0001 and cid0010: inf0003
...: ...
...: ...

  For example, suppose that there is an advertiser who wants to advertise jewelry to a user group related to a cluster in which the homepage (URL) of “Concert Hall A” is included. In this case, first, as the action specifying information as shown in Table 3, the URL of the homepage of “Concert Hall A” may be set. Further, as the provided information ID paired with this, the URL of the Web page on which the advertisement is posted may be set.

  Next, an operation example of the information processing apparatus 100 in the present embodiment will be described.

  First, the cluster processing unit 104 will be described. First, the cluster processing unit 104 generates a cluster from the action specifying information stored in the action history storage unit 103.

  A cluster generation (clustering) method will be described. In the following description, access by the user terminal device by the user's operation is abbreviated as user access. There are various methods for clustering, as is well known. For example, access histories for a predetermined period are compared among users (user IDs), changes in access destinations (behavior identification information) are compared, users who have similar access changes are identified, There is a way to group.

  For example, the user 1 accesses each Web page “Web page A → Web page B → Web page C” and the user 2 accesses each Web page “Web page E → Web page F” in the action history storage unit 103. Consider a case where a state is accessed in which the user 3 accesses each Web page in the order of “Web page A → Web page B → Web page D”. If the user's action history is compared, and a user whose transitions (transitions taking time into account) of two Web pages (behavior specific information) match as a threshold is set as similar, the user 1 And the user 3 become similar users, and the web page A, the web page B, the web page C, and the web page D can be made into one cluster in the behavior specifying information of these users.

  For example, if a cluster including the Web page D is selected, the cluster is taken out, and the user 1 and the user 3 associated with the taken-out cluster are information provision targets. It becomes. At this time, the user 1 is not accessing the Web page D, but is an object of information provision. This is because “user 1 is similar to user 3 who has accessed web page D, so even if he / she has not accessed web page D, he / she is a target for providing information related to web page D”. Become.

  In the above-described identification (classification) of similar users, the probability of accessing each Web page is obtained from the action history, and the URL (behavior specifying information) that is information for identifying the Web page with a high probability is similar. A method of classifying (clustering) existing users as similar users may be used. These clustering methods may use other techniques.

  Further, clustering may be performed not only based on the similarity of the transition (the state in which the order is taken into consideration) with respect to the web page, but based on the similarity of the accessed web page. For example, a user 1 who accesses “Web page A, Web page B, Web page C”, a user 2 who accesses “Web page E, Web page F”, and “Web page A, Web page B, Web page D”. , The user 1 and the user 3 who accessed the web page A and the web page B are classified as similar, and the actions of the similar (classified) user 1 and the user 3 are classified. One cluster may be generated from Web page A, Web page B, Web page C, and Web page D, which are behavior specifying information included in the history.

  In the identification (classification) of similar users described above, the behavior history of each user stored in the behavior history storage unit 103 can be processed and used as follows when generating a cluster. For example, the Markov process is used to determine the state transition probability of access to each user's action specifying information (URL) from the action history of each user as shown in Table 1 above. Here, the state transition probability is a probability for the action specifying information (behavior specifying information to be changed) to be accessed after the user accesses certain action specifying information (URL).

  As shown in Table 3, in the action history, action specifying information (latest action history) last accessed by each user is shown. For example, the user “uid0001” is “cid0005” and the user “uid0002” is “cid0001” as the action specifying information last accessed by each user. The probability of which action specifying information is accessed next to the action specifying information can be obtained by a Markov process. In this case, for example, “uid0001” last accessed “cid0005”, but after that, the probability of accessing “cid0005” or “cid0015” is 0.3%, and the probability of accessing “cid0010” is 0. 29% and so on. Using these probabilities, it is possible to determine that users having similar probabilities for each action specifying information are similar, and generate a cluster.

  For example, as shown in Table 4 below, “uid0001” and “uid0003” all have similar probabilities with respect to “cid0005”, “cid0010”, and “cid0015”. Therefore, it can be determined that these users are similar to each other. In this case, as shown in Table 2 above, the cluster includes the action specifying information and each user ID of each user recorded in the action history.

Table 4
Action specific information user ID cid0001 cid0005 cid0010 cid0015…
uid0001 0 0.3 0.29 0.3…
uid0002 0.4 0 0.03 0.3…
uid0003 0 0.3 0.35 0.3…
uid0004 0 0 0.01 0.02…
::::: ...

  For example, when the action specifying information is a web page (URL), a user who is interested in a car frequently accesses a car-related web page, and repeatedly accesses a car-related web page. It is estimated that In such a case, the transition of access to these Web pages may be looped. Users with similar loops can be a single cluster.

  By the way, when the behavior specifying information is a product that is traded using the Internet, in the case of product sales, the same product is rarely purchased, and the transition due to the purchase of the same product is rare. However, when the product to be traded is a so-called consumable product, it is assumed that the same product (consumable product) is purchased in a certain chain after a certain period of time. In such a case, the transition of purchasing consumables is a loop, and users with similar loops can be made into one cluster.

  In addition, the cluster may include all the action specifying information included in the action histories of a plurality of users classified as similar as a cluster component, and only the action specifying information selected under a predetermined condition is clustered. It may be added as an ingredient. As the predetermined condition, for example, time information recorded (stored) in the action history together with the action specifying information from all the action specifying information included in the action history of the plurality of classified users is from a certain time. This is a condition for selecting a new one (with a predetermined time zone) and making it a cluster (leaving it as a component of the cluster). In this case, in Table 1 described above, the action specifying information recorded with the “access month / date” whose “access month / date” is newer than “20060101” is left as a cluster component. By doing in this way, the cluster of the state according to the present user's preference comes to be obtained, and since action specific information contained in a cluster can be reduced, storage capacity can be controlled.

  In addition, for example, a cluster is generated from action identification information that is equal to or more than a predetermined number of overlaps over the action history of each classified user among all the action identification information included in the action history of a plurality of classified users. Also good. By doing in this way, it can be in the state where the similarity of the user was reflected more in the generated cluster, and the action specifying information included in the cluster can be reduced, so that the storage capacity can be suppressed. it can. In this case, a cluster may be generated by selecting the action specifying information in consideration of the temporal transition similarity of the action specifying information in each action history.

  Thus, the user who the state which changes and closes in the specific action specific information obtained in this way can be made into a similar user. In this way, a set of action specifying information included in the action history of similar users can be made into one cluster.

  As described above, when a cluster is generated, the cluster processing unit 104 stores the generated cluster in the cluster storage unit 105. Here, when the action history including the various action specifying information described above is stored in, for example, the action history storage unit 103, the above-described cluster generation process may be executed. Further, the behavior generation history may be periodically sent to the cluster processing unit 104 so that the above-described cluster generation processing may be realized.

  Next, a process of providing information such as selecting a target user and making an advertisement by using a cluster stored in the cluster storage unit 105 will be described with reference to the flowchart of FIG. First, for example, any action specifying information stored in the provided information storage unit 106 is selected by an operation of an administrator who manages the information processing apparatus 100, or the collation processing unit 107 is periodically or designated. When the information stored in the provided information storage unit 106 is read out, the matching processing unit 107 searches the cluster storage unit 105 using the selected action specifying information as a key (step S201). Here, if there is a cluster including the action specifying information corresponding to the cluster storage unit 105 (step S202), the collation processing unit 107 extracts the user ID included in the corresponding cluster (step S203).

  Next, the information provision control part 108 takes out provision information ID matched with the selected action specific information mentioned above from the provision information storage part 106 (step S204). Next, the information provision control unit 108 generates information to be provided from the extracted provision information ID (step S205), and the generated information is sent to the user terminal device 131 corresponding to the user ID extracted by the verification processing unit 107. It is sent out (step S206). Here, the provided information ID may be extracted at the cluster search stage (initial stage). In step 302, if there is no cluster including the corresponding action specifying information, it is determined that there is no user to be provided with information, and the process ends.

  In the above-described search using the action specifying information as a key, for example, if the action specifying information is “cid0005” shown in Table 3 and the corresponding provided information ID is “inf0001”, the matching processing unit 107 first displays “ The cluster storage unit 105 is searched for a cluster including “cid0005”. When each cluster is stored in the cluster storage unit 105 as shown in Table 2, cluster 1 and cluster 2 in Table 2 are selected by the above search. Further, the collation processing unit 107 extracts “uid0001”, “uid0003”, and “uid0012” included in the cluster 1 and the cluster 2. Further, the information provision control unit 108 sends (notifies) information identified by “inf0001”, for example, a URL, to the user terminal device 131 corresponding to the extracted “uid0001”, “uid0003”, “uid0012”. Or presented (displayed) on a Web page accessed by the user terminal device 131.

  For example, in the case of an advertiser who wants to advertise jewelry as described above, “cid0005” is the URL of the homepage of concert hall A. The URL to be sent is, for example, an address of a web page where an advertisement prepared by the advertiser on the information provider terminal 140 or the service providing server 150 can be browsed, and a jewelry advertisement is posted on the web page. Then, the advertisement of jewelry is made to the user.

  This information is transmitted when, for example, the user terminal device 131 of the target user ID is accessing a Web page provided on a portal site that operates the information processing apparatus 100. May be displayed in a state that is visible to the user who is browsing. Further, instead of displaying on the web page being browsed, a new window may be developed to display the web page with the corresponding URL. Further, when accessing a Web page or the like related to information to be sent, the information may be sent and the corresponding URL may be displayed so as to be visible to the user. The information may be notified as an email to the email address of the user identified by the corresponding user ID.

  Here, in the action history stored in the action history storage unit 103, there is an access history for the URL to be transmitted by any one of the extracted "uid0001", "uid0003", and "uid0012". In this case, information (URL notification) may not be provided to a user who has an access history. In other words, a user having behavior identification information including information to be provided in the behavior history is excluded from information provision targets. By doing so, provision of duplicate information can be suppressed.

  Whether or not a user who has an access history for the URL identified by the provided information ID should again browse the Web page of the URL may be set in advance. Moreover, you may make it attach the flag of whether duplication is permitted to the group of provision information ID and action specific information memorize | stored in the provision information storage part 106. FIG. In addition, conditions for permitting duplication (users who are not used within a certain period of time, users who are used three times or less in the history, etc.) are described as criteria for judgment, and based on this description Information provision operation may be controlled. When conditions are described in this way, a process for confirming the described conditions may be provided to notify the user of information that meets the described conditions.

  As described above, according to the present embodiment, it is possible to provide information to a suitable user layer only by specifying the action specifying information and the provided information ID. Further, in the above description, information is provided to the user terminal device 131 and the mobile terminal device 132 that can be directly connected to the information processing device 100 of the portal operator. However, the present invention is not limited to this. . For example, it goes without saying that information may be provided to the user terminal device 133 that can be connected to the information processing apparatus 100 via the Internet 120.

  For example, the action specifying information and the provided information ID are selected and set by the advertiser, for example, but the action specifying information is a component of a certain feature in the cluster, and is not necessarily directly accessed by the user himself / herself. The personal attribute of the user is not recognized by the advertiser (information provider).

  As described above, by analyzing various information (behavior identification information) managed by service providers such as portal providers and ISPs, the characteristics of each user are analyzed, and information that emphasizes the characteristics is recommended. Is possible. The user characteristics indicate a user's living environment and preferences, strengths and preferences of preferences, timing for executing something, amount of information to be referred to when necessary, relationship with other users, and the like. The recommended information includes advertisements as well as search results and service information. If the advertisement is necessary information for the user, it is no longer an advertisement but useful information.

  As described above, portal service providers and ISPs and other service providers enter between service providers such as users and advertisers, which not only protects user personal information but also accompanies searches. Such advertisements are not distributed in the case of information that is not necessary for the corresponding user because the portal company optimizes the information for the corresponding user when providing the search result. Accordingly, it is possible to minimize the distribution of unnecessary information to the user. The above-described embodiment is an example of optimizing advertisement in such a service as a portal company. Information provision in the present embodiment can be applied to optimization of advertisements provided to users in services provided by service providers such as portal providers, advertisers, and users (portal users).

  The timing of providing information, such as performing this advertisement, is partially displayed on the display unit of the user terminal device 131 when the user 130 logs in to the portal operator (information processing apparatus 100) and accesses a customized user-specific page. It may be displayed in the above window, or may be transmitted by regular delivery e-mail or the like. Further, an advertisement may be displayed at the time of accessing the URL identified by the other action specifying information included in the cluster including the action specifying information and when executing the search. For example, if the URL is included in the search result returned from the search engine when the search is executed, the position in the search result list close to the search result indicating the URL Then, the URL of the advertisement may be displayed. In addition, when the search result does not include the URL, but includes another URL identified by other action specifying information included in the cluster to which the action specifying information belongs, In the list of results, the URL of the advertisement may be displayed at a position close to the search result including other URLs.

  Information such as advertising to a so-called blog site established by the user 130 may also be provided. Information may be provided in response to an explicit request from the user 130. For example, an information providing button is prepared on the screen of a website that the user 130 is browsing, and information is provided when the button is selected (clicked) by operating the mouse. Further, the information may be provided not by providing information via the Internet but by mail of a document or article such as direct mail.

  In the above description, the action specifying information and the provided information ID are described in the case of the URL. However, the action specifying information is made to correspond to the product purchased through online shopping, and the provided information ID is made to correspond to the attached product of the purchased product. Good. Further, as the provided information ID, a URL of a Web page including information on the attached product may be associated. In addition, when the current location can be grasped by a mobile phone using a so-called GPS technology, a place name or the like is set as action specifying information, and a URL for a store advertisement related to the place name is set in the provided information ID. May be set. Information may be provided when a user performs a predetermined search or the like by accessing the mobile phone to the Internet or when position information is transmitted. In this case, it is better to update the cluster by performing the clustering of the cluster processing unit at a high frequency in accordance with the transition of the region.

  The above-mentioned provision of information is recorded when an information provider such as an advertiser registers data, when a matching process is performed at a preset timing in the system, or when any access or action from the user occurs. It may be performed after it is determined that there is information that matches the user, such as the time point. This timing may be in accordance with the providing method selected by the user himself.

  In addition, priority may be given to the various information provision timings described above, and information provision may be performed from timings with high priority. For example, when the user does not access a predetermined website during the set time period, the user waits for position information to be sent or execution of a search, and these are not received within the predetermined time. Waits for acceptance of an explicit request from the user, and if this is also not done within a predetermined time, informs the user's email address of the information as an email. At the same time, any of the above-described information providing methods may be set with a predetermined ranking and executed (information provision) at the ranking timing.

  As described above, the provision of information is established by the user specified by the user ID, such as sending an e-mail, displaying on a web page provided at a site accessed by the user terminal device, as described above. It may be performed by means selected from a plurality of means including display on a Web page provided at a certain site. Information may be provided by a selected means in a preset order.

  Note that the information processing apparatus 100 illustrated in FIG. 1 is a computer device including a CPU, a main storage device, an external storage device, a network connection device, and the like, and the CPU is operated by a program developed in the main storage device. Each function described above is realized. Each storage unit and each function may be distributed to a plurality of computer devices connected to a network including the Internet.

It is a block diagram which shows the structure of the information processing apparatus in embodiment of this invention. 5 is a flowchart for explaining an operation example of the information processing apparatus 100.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 100 ... Information processing apparatus, 101 ... User information storage part, 102 ... Action history physiology part, 103 ... Action history storage part, 104 ... Cluster processing part, 105 ... Cluster storage part, 106 ... Provision information storage part, 107 ... Collation process , 108 ... Information provision control part, 109 ... Transmission / reception part, 120 ... Internet, 130 ... User, 131 ... User terminal device, 132 ... Mobile terminal device, 133 ... User terminal device, 140 ... Information provider terminal.

Claims (10)

An action history storage unit in which action identification information obtained from a state in which the terminal device is connected to the network is stored as an action history together with a user ID for identifying the user;
By classifying user IDs having similar behavior histories stored in the behavior history storage unit, the user IDs are generated in association with the classified user IDs of clusters including behavior identification information included in the behavior histories of the classified user IDs. A cluster processing unit;
A cluster processing unit that selects a cluster including the set action specifying information from among the clusters generated by the cluster processing unit, and includes at least a matching processing unit that extracts a user ID associated with the selected cluster as an information providing target; An information processing apparatus characterized by the above. The information processing apparatus according to claim 1,
The cluster processing unit classifies user IDs with similar temporal transitions of behavior specifying information included in the behavior history. The information processing apparatus according to claim 1 or 2,
In the behavior history storage unit, the behavior identification information is stored as the behavior history together with time information in addition to the user ID,
The cluster processing unit generates the cluster from action specifying information in which the time information is in a predetermined time zone. The information processing apparatus according to claim 1 or 2,
The cluster processing unit generates the cluster from the action specifying information having a predetermined number of overlaps or more over the action history of each classified user ID from the action specifying information included in the action history of the classified user ID. An information processing apparatus characterized by the above. The information processing apparatus according to any one of claims 1 to 4,
The information processing apparatus, wherein the collation processing unit excludes a user who has at least a part of information to be provided in an action history from an information provision target. The information processing apparatus according to any one of claims 1 to 5,
Sending out an e-mail to the user terminal device corresponding to the user ID extracted by the collation processing unit as the information providing target, the information associated with the action specifying information used for selecting the cluster, the user terminal From a plurality of means including a display on a Web page provided on a site accessed by the apparatus and a display on a Web page provided on a site established by the user specified by the user ID An information processing apparatus comprising: an information provision control unit provided by a selected means. The information processing apparatus according to claim 6.
The information provision control unit provides the information by the selected means in a preset order. An information providing system connected to an information provider terminal and a user terminal and providing information registered by the information provider terminal to the user terminal,
The information provider includes a plurality of clusters that are configured of behavior specifying information acquired by accessing the user terminal and reflect the similarity of the behavior history associated with the user terminal device or the user identification information. Means for selecting a cluster including action specifying information registered together with the information from the terminal, and selecting a user terminal associated with the selected cluster;
An information providing system comprising: at least notification means for notifying the selected user terminal of the information. By classifying user IDs that have similar behavior histories that include behavior identification information that is configured from information obtained from the state in which the terminal device is connected to the network and that indicates the state of use of the network by the user, Generate a cluster of behavior identification information included in the behavior history in association with the classified user ID,
From the clusters generated by this cluster processing unit, select a cluster including the set action specifying information, retrieve the user ID associated with the selected cluster, and provide the user ID retrieved by the matching processing unit An information processing method characterized by being targeted. On the computer,
An action history storage function in which action specifying information that is configured from information obtained from a state in which the terminal device is connected to the network and indicates a state of use of the network by the user is stored as an action history together with the user ID for identifying the user;
By classifying user IDs with similar behavior histories stored by the behavior history storage function, a cluster of behavior identification information included in the behavior history of the classified user ID is generated in association with the classified user ID. Cluster generation processing function,
A program for realizing a matching processing function that selects a cluster including set action specifying information from clusters generated by the cluster generation processing function and extracts a user ID associated with the selected cluster.

Images