JP2009266056A - Information recommendation device, information recommendation method, program and recording medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To achieve personalization in taking the versatility of a user into consideration, even when a user does not sufficiently act as a persona. <P>SOLUTION: An information recommendation device stores a persona cluster database, in which persona cluster identifiers respectively identifying a plurality of clusters having classified a plurality of distribution information pieces for each field are associated with first distribution information identifiers respectively identifying the distribution information pieces belonging to the clusters respectively corresponding to the persona cluster identifiers, in a storage section. The information recommendation device receives a set of second distribution information identifiers respectively identifying the plurality of distribution information pieces from a terminal device, obtains similarities between a set of the first distribution information identifiers associated with the same persona cluster identifier in the persona cluster database and a set of the second distribution information identifiers, for each persona cluster identifier, selects one or more persona cluster identifiers in the descending order of the similarities, and extracts the selected persona cluster identifiers to transmit to the terminal device. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention provides information for recommending a persona, which is a different personality to be used for each field of distribution information, to a terminal device to which information is distributed through the network when the user receives distribution of information through the network. It relates to the technology to be provided.

  There is a service that searches distribution information through the network and recommends it to users. Recommended distribution information includes moving images, still images, music, product information, news information, blog information, telephone number information, service information, and the like. In addition, there is a personalized technology as a technology for adapting these services to the user's hobbies and preferences. This technology adapts the above services to the user at the granularity of the individual user by utilizing the personal information and action history of the user. One example is a service that recommends distribution information to a user based on a user's action history on the network. Collaborative filtering (see, for example, Non-Patent Document 1) and ID management technology are basic technologies for realizing the personalized technology. Collaborative filtering is a methodology that accumulates preference information of many users and automatically makes inferences using information of other users with similar preferences to a certain user. The ID management technique is a technique for assigning a unique ID to each user and performing user authentication, personal information management, and the like using the ID. The service provider can adapt the above service to the user based on the personal information of the user collected by the ID management technology.

  On the other hand, users have multiple aspects (personality). One example is a case in which a single user has a plurality of aspects such as a programmer aspect, a railway enthusiast aspect, a musician aspect, and a 30-year aspect. Each side is called a persona. Under ID management technology, users can control privacy by using different IDs for each persona. One example is the proper use of a plurality of E-MAIL addresses by the user. There is a method disclosed in Patent Document 1 as a method for realizing a personalized technique with a persona granularity instead of a user granularity. In patent document 1, the personalized technique which considered the multiplicity of the above users is implement | achieved using the profile of a persona. An example of a persona profile includes attribute information such as a persona nickname, personal information, contents and services used, and information transmitted in the name of the persona.

In such a personalized technology considering the multiplicity of users, the trouble of creating a profile for each persona becomes an issue. As a technique for automatically generating a profile for each persona, there is one disclosed in Patent Document 2. In this technique, a persona profile is automatically generated based on a persona action history such as a service applied via a site or network accessed by the persona.
Paul Resnick, Neophytos Iacovou, Mitesh Suchak, Peter Bergstrom, John Riedl, “GroupLens: An Open Architecture for Collaborative Filtering of Netnews”, [online], [Search March 26, 2008], Internet <http: // www .si.umich.edu / ~ presnick / papers / cscw94 / GroupLens.htm> JP 2005-524132 A JP 2000-66981 A

  However, since the method disclosed in Patent Document 2 automatically generates a profile based on the action history of the persona, the persona profile cannot be appropriately created unless the user acts sufficiently as a persona. The persona cannot be matched to the user side. In this case, personalization in consideration of the user's versatility cannot be realized. This is a barrier to starting a new persona.

  The present invention has been made in view of such points, and a technology capable of realizing personalization in consideration of the multifaceted nature of the user even in a situation where the user is not sufficiently acting as a persona. The purpose is to provide.

  In the first aspect of the present invention, in order to solve the above-mentioned problem, when the information recommendation apparatus receives a distribution of information through the network, the information recommendation apparatus receives each distribution information field. Send information to recommend personas, which are different personalities. For this purpose, the information recommendation device firstly identifies each persona cluster identifier that identifies a plurality of clusters obtained by classifying a plurality of pieces of distribution information that can be distributed through a network, and each person belonging to a cluster corresponding to the persona cluster identifier. A persona cluster database in which the first distribution information identifiers for identifying the distribution information are associated with each other is stored in the storage unit. Then, the information recommendation device receives a second set of distribution information identifiers for identifying a plurality of pieces of distribution information from the terminal device and associates each persona cluster identifier with the same persona cluster identifier in the persona cluster database. The similarity between the first set of distribution information identifiers and the second set of distribution information identifiers is obtained, one or more persona cluster identifiers are selected in descending order of the similarity, and the selected persona cluster identifiers are extracted. , It is transmitted to the terminal device.

  Here, it can be presumed that the second set of distribution information identifiers received by the information recommendation device from the terminal device indicates hobbies and preferences at the granularity of the individual user of the terminal device. In addition, selecting one or more persona cluster identifiers in descending order of similarity to each of the fields that are likely to match the user's hobbies and preferences, that is, each side that matches each side of the user. This is equivalent to selecting a field. Then, transmitting each selected persona cluster identifier to the terminal device corresponds to transmitting the ID of the cluster to which the persona that matches each side of the user belongs. With this ID management technique using the persona cluster identifier, personalization that takes into account the user's versatility can be realized even in a situation where the user is not sufficiently acting as a persona.

  In the second aspect of the present invention, first, the information recommendation device includes a user identifier for identifying each user and a cluster to which a distribution information identifier for identifying distribution information distributed to the terminal device of the user belongs. A persona database in which persona cluster identifiers to be identified are associated with each other is stored in the storage unit. The information recommendation device receives a set of persona cluster identifiers transmitted from the terminal device, and for each user identifier, a set of persona cluster identifiers associated with the same user identifier in the persona database, and a receiving unit The similarity with each received persona cluster identifier set is obtained, and one or more user identifiers are extracted in descending order of similarity. Then, the information recommendation device extracts each persona cluster identifier associated with the extracted user identifier in the persona database, excluding each received persona cluster identifier, and outputs it to the terminal Send to device.

  Here, it can be estimated that the persona cluster identifier received by the information recommendation device from the terminal device corresponds to a persona that matches at least some aspects of the user of the terminal device. In addition, extracting one or more user identifiers having a high degree of similarity with the set corresponds to extracting IDs of other users whose at least some aspects coincide with the user of the terminal device. . In addition, each persona cluster identifier associated with the extracted user identifier, and each persona cluster identifier excluding each received persona cluster identifier is extracted and transmitted to the terminal device. This corresponds to transmitting the ID of the cluster to which the persona that is likely to match the other side of the user belongs to the terminal device. With this ID management technique using the persona cluster identifier, personalization that takes into account the user's versatility can be realized even in a situation where the user is not sufficiently acting as a persona.

  As described above, in the present invention, it is possible to realize personalization in consideration of the multiplicity of users even when the user is not sufficiently acting as a persona.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
[First Embodiment]
First, a first embodiment of the present invention will be described.
<Configuration>
FIG. 1 is a conceptual diagram illustrating the configuration of an information recommendation system 1 according to the first embodiment.
As shown in FIG. 1, the information recommendation system 1 of this embodiment includes an information recommendation device 10, terminal devices 20-n (n = 1,..., N, N is an integer equal to or greater than 1), and an information distribution device 30. −m (m = 1,..., M, M is an integer of 1 or more), and the information recommendation device 10, the terminal device 20-n, and the information distribution device 30-m can communicate with the network 40. It is configured as follows. In addition, the information recommendation device 10, the terminal device 20-n, and the information distribution device 30-m are publicly known having a communication function including a CPU (Central Processing Unit), a RAM (Random Access Memory), a communication unit, and the like. It is an apparatus configured by reading a predetermined program into a computer. For simplification of explanation, the present embodiment shows an example in which only one information recommendation device 10 exists, but two or more information recommendation devices 10 may exist.

  FIG. 2 is a block diagram illustrating a functional configuration of the information recommendation device 10 according to the first embodiment. FIG. 3 is a block diagram illustrating an example of a functional configuration of the information similarity determination unit 13b.

  As shown in FIG. 2, the information recommendation device 10 of this embodiment includes a persona cluster database storage unit 11a, a persona database storage unit 11b, a temporary storage unit 11c, a transmission unit 12a, a reception unit 12b, a control unit 13a, and information similarity determination. A unit 13b and a persona identifier assigning unit 13c. As illustrated in FIG. 3, the information similarity determination unit 13b includes, for example, a count unit 13ba and a comparison unit 13bb. Here, examples of the persona cluster database storage unit 11a, the persona database storage unit 11b, and the temporary storage unit 11c are a hard disk device, a RAM, a cache memory, a register, or a storage area that combines at least a part thereof. In addition, examples of the transmission unit 12a and the reception unit 12b are communication units such as modems that are driven under the control of the CPU loaded with a predetermined program. Moreover, the example of the control part 13a, the information similarity determination part 13b, and the persona identifier assignment part 13c is CPU which the predetermined | prescribed program was read. The control unit 13a controls the entire processing of the information recommendation device 10. Although not described below, the temporary storage unit 11c stores data output from each function unit as necessary, and each stored data is read from other function units as necessary. . Also, sending data from a certain function unit to another function unit means, for example, that data output from a certain function unit is stored in the temporary storage unit 11c, and the other function unit is transferred from the temporary storage unit 11c to the data. Is to read.

  FIG. 4 is a block diagram illustrating a functional configuration of the terminal device 20-n according to the first embodiment. Since each terminal device 20-n has the same configuration, one terminal device 20-n will be described as the terminal device 20 below.

  As illustrated in FIG. 4, the terminal device 20 according to the present embodiment includes a user identifier storage unit 21a, a distribution information identifier database storage unit 21b, a temporary storage unit 21c, a transmission unit 22a, a reception unit 22b, a control unit 23a, and response information. A generation unit 23b, a writing unit 23c, a display unit 24a, and an input unit 24b are included. Here, examples of the user identifier storage unit 21a, the distribution information identifier database storage unit 21b, and the temporary storage unit 21c are a hard disk device, a RAM, a cache memory, a register, or a storage area that combines at least a part thereof. . In addition, examples of the transmission unit 22a and the reception unit 22b are communication units such as modems that are driven under the control of the CPU loaded with a predetermined program. An example of the control unit 23a and the response information generation unit 23b is a CPU in which a predetermined program is read. The example of the writing unit 23c is a CPU that reads a predetermined program or a reader / writer that is driven under the control of the CPU. An example of the display unit 24a is a display unit including a video card and a display. Examples of the input unit 24b are a keyboard and a mouse. The control unit 23a controls processing of the entire terminal device 20 including the control unit 23a. Although not described below, the temporary storage unit 21c stores data output from each function unit as necessary, and the stored data is read from other function units as necessary. . Also, sending data from a certain function unit to another function unit means, for example, that data output from a certain function unit is stored in the temporary storage unit 21c, and other function units are transferred from the temporary storage unit 21c to the relevant data. Is to read.

  The information distribution device 30-m is a device that distributes various types of distribution information. Examples thereof include a video distribution device, a still image distribution device, a music distribution device, a product information distribution device, a news information distribution device, a blog information distribution device, A telephone number information distribution device, an online bank device, an online transaction device, and the like. For the sake of simplicity, each information distribution device 30-m is hereinafter referred to as an information distribution device 30.

<Processing>
Next, the processing of this embodiment will be described.
[Preprocessing]
The preprocessing that is the premise of the processing of this embodiment will be described.
First, in the persona cluster database storage unit 11a of the information recommendation device 10, each persona cluster identifier (CID) that identifies each of a plurality of clusters obtained by classifying a plurality of pieces of distribution information that can be distributed through the network 40 by each information distribution device 30. ) And a first distribution information identifier (DID1) for identifying each distribution information belonging to a cluster corresponding to the persona cluster identifier (CID), respectively, is stored. For example, the information recommendation device 10 obtains a plurality of distribution information from each information distribution device 30 through the network 40, clusters them for each field, assigns a persona cluster identifier to each cluster, This is performed by associating the persona cluster identifier with the first distribution information identifier for identifying the corresponding distribution information. Here, clustering of a plurality of pieces of distribution information may be performed by a person referring to each piece of distribution information, or automatically using a known clustering method such as a k-means method or hierarchical clustering. . When clustering is automatically performed, for example, each piece of distribution information is expressed as a document vector, and clustering is performed based on the similarity between the document vectors. A known method may be used to generate the document vector (for example, “Kazuaki Kishida,“ Document Clustering Technique: Literature Review ”, Mita Library and Information Society, N0.49 (2003), p. 33- 75 ”,“ Takenobu Tokunaga, Junichi Sakurai, “Information Retrieval and Language Processing Language and Computation”, The University of Tokyo Press, 1999 ”,“ Ken Kenji, Kazuhiko Tsuda, Masato Sasabori, “Information Retrieval Algorithm”, See Kyoritsu Publishing, 2002 ”). Examples of persona cluster identifiers are codes and numbers unique to each cluster, and examples of distribution information identifiers are URLs (Uniform Resource Locators) and IDs corresponding to distribution information.

  FIG. 5A is a diagram illustrating a data configuration of the persona cluster database stored in the persona cluster database storage unit 11a. FIG. 5A shows an example in which the distribution information identifier is a URL and the persona cluster identifier is a number unique to each cluster.

  In the persona database storage unit 11b of the information recommendation device 10, a user identifier (UID) for identifying each user at the granularity of the individual user and a user terminal device corresponding to the user identifier (UID) are stored. The persona cluster identifier (CID) of the cluster to which the delivery information identifier of the delivered delivery information belongs is associated with the persona identifier (PID) that has already been issued to the user corresponding to the user identifier (UID). Store the persona database. The persona identifier (PID) is an identifier for identifying each persona with the granularity of the persona. As will be described later, each persona identifier (PID) issued in this embodiment is associated with a persona cluster identifier (CID). That is, each persona identifier (PID) is associated with one of the fields of distribution information. The persona means a different personality to be used for each field of distribution information when the user receives distribution of information through the network 40 to the terminal device 20 through which information is distributed through the network 40.

  FIG. 5B is a diagram illustrating a data configuration of the persona database stored in the persona database storage unit 11b. In the example of FIG. 5B, each user identifier is associated with one or more persona identifiers, and each persona cluster identifier is associated with one or more persona identifiers.

  On the other hand, a user identifier (UID) for identifying a user who uses the terminal device 20 is stored in the user identifier storage unit 21 a of each terminal device 20. In addition, each user of each terminal device 20 accesses each information distribution device 30 using the terminal device 20 at the granularity of the individual user without being conscious of the persona. It is assumed that the delivery has been received. The distribution information identifier database storage unit 21b of each terminal device 20 stores a second distribution information identifier (DID2) for identifying each distribution information received by the terminal device 20.

  FIG. 6A is a diagram illustrating a data configuration of a distribution information identifier database stored in the user identifier storage unit 21a of the terminal device 20 to which no persona identifier is assigned. In the example of FIG. 6A, a plurality of second distribution information identifiers that are URLs are stored as elements of the distribution information identifier database. The second distribution information identifiers in the distribution information identifier database at this stage are not classified for each field, and second distribution information identifiers corresponding to the distribution information in a plurality of fields are mixed.

[Persona recommendation processing]
7 and 8 are flowcharts for explaining the persona recommendation process of the first embodiment. Hereinafter, the persona recommendation processing of this embodiment will be described with reference to these drawings.
First, a set of second distribution information identifiers (DID2) for identifying a plurality of pieces of distribution information stored in the distribution information identifier database storage unit 21b of the terminal device 20 is read and sent to the transmission unit 22a. The set of second distribution information identifiers (DID2) may be a set having all the second distribution information identifiers (DID2) stored in the distribution information identifier database storage unit 21b as elements. The second distribution information identifier (DID2) arbitrarily selected by the user of the terminal device 20 may be a set having the second distribution information identifier (DID2) not associated with the persona identifier as an element. It may be a set whose element is. The transmission unit 22a transmits the set of second distribution information identifiers (DID2) to the information recommendation device 10 via the network 40 (step S1).

  This set of second distribution information identifiers (DID2) is received by the receiving unit 12b of the information recommendation device 10 and sent to the information similarity determination unit 13b (step S2). The information similarity determination unit 13b refers to the persona cluster database stored in the persona cluster database storage unit 11a, and associates each persona cluster identifier (CID) with the same persona cluster identifier (CID) in the persona cluster database. The similarity between the set of first distribution information identifiers (DID1) and the set of sent second distribution information identifiers (DID2) is obtained, and one or more persona cluster identifiers (CIDs) in descending order of the similarity ) Is selected, and the selected persona cluster identifier (CID) is extracted (step S3).

<< Specific Example of Step S3 >>
In this example, a set of second distribution information identifiers (DID2) is input to the count unit 13ba of the information similarity determination unit 13b, and the count unit 13ba includes a set of second distribution information identifiers (DID2) and a persona cluster. The second distribution information identifier (DID2) and the first distribution information identifier (DID1) are compared with a set of first distribution information identifiers (DID1) associated with the same persona cluster identifier (CID) in the database. ) Is counted. The count unit 13ba sets the count number (CT) as the similarity between the set of the second distribution information identifier (DID2) and the cluster corresponding to the persona cluster identifier (CID), and the count number (CT), The persona cluster identifier (CID) and the second distribution information identifier (DID2sub) that matches the first distribution information identifier (DID1) are stored in the temporary storage unit 11c in association with each other. The counting unit 13ba performs the same process for each persona cluster identifier (CID) in the persona cluster database. The comparison unit 13bb compares the count numbers (CT) respectively associated with the persona cluster identifiers (CID) stored in the temporary storage unit 11c, and the count number (CT) is large (similarity is high). ) One or more persona cluster identifiers (CID) are selected in order, and each selected persona cluster identifier (CID) and a corresponding second distribution information identifier (DID2sub) are output.

  It should be noted that the number of coincidence between the second distribution information identifier (DID2) and the first distribution information identifier (DID1) is simply counted, and the count number (CT) is not used as the degree of similarity. A weight may be set for the distribution information identifier (DID1), and the count number (CT) considering the weight may be used as the similarity. Specifically, the count unit 13ba is set to the matched first distribution information identifier (DID1) when the second distribution information identifier (DID2) and the first distribution information identifier (DID1) match. You may count up by the weight that you have. The number of persona cluster identifiers (CID) extracted by the comparison unit 13bb may be set in advance, or a threshold value related to the count number (CT) is set, and the count number (CT) equal to or greater than the threshold value is set. All the persona cluster identifiers (CIDs) that have become may be extracted (end of description of << specific example of step S3 >>).

  The persona cluster identifier (CID) extracted in step S3 and the second distribution information identifier (DID2sub) that matches the first distribution information identifier (DID1) associated with the persona cluster identifier (CID). Are associated with each other and sent to the transmitter 12a. The transmitter 12a transmits these pieces of information to the terminal device 20 via the network 40 (step S4).

  These pieces of information are received by the receiving unit 22b of the terminal device 20 (step S5), and first sent to the display unit 24a. The display unit 24a displays a set of the persona cluster identifier (CID) and the second distribution information identifier (DID2sub), and also displays information for prompting the user to select a persona cluster identifier (CID). (Step S6). The input unit 24b receives input of selection information (SI) for specifying the selected persona cluster identifier (CID), and sends the selection information (SI) to the response information generation unit 23b (step S7). The response information generating unit 23b further includes a persona cluster identifier (CID) and a second distribution information identifier (DID2sub) received by the receiving unit 22b, and a user identifier (UID) read from the user identifier storage unit 21a. ) And. The response information generation unit 23b generates response information including a user identifier (UID), a persona cluster identifier (CID), a second distribution information identifier (DID2sub), and selection information (SI), and transmits the response information to the transmission unit 22a. (Step S8). Also in the response information, each persona cluster identifier (CID) is associated with a corresponding second distribution information identifier (DID2sub). The transmission unit 22a transmits the response information to the information recommendation device 10 via the network (step S9).

  The response information is received by the receiving unit 12b of the information recommendation device 10 (step S10) and sent to the persona identifier assigning unit 13c. The persona identifier assigning unit 13c issues new persona identifiers (PIDs) as many as the number of persona cluster identifiers (CIDsel) indicated by the selection information (SI) included in the response information (step S11). At this time, the persona identifier assigning unit 13c refers to the persona database stored in the persona database storage unit 11b and issues a persona identifier (PID) that does not overlap with the issued persona identifier. Next, the persona identifier assigning unit 13c associates each issued persona identifier (PID) with each persona cluster identifier (CIDsel) indicated by the selection information (SI) on a one-to-one basis, and further includes response information. The data is stored in the persona database storage unit 11b in association with the user identifier (UID). Thereby, data of the persona database is added. Each issued persona identifier (PID), each persona cluster identifier (CIDsel) associated therewith, and a second distribution information identifier (DID2sub) corresponding to each persona cluster identifier (CIDsel) Persona information associated with is sent to the transmitter 12a. The transmission unit 12 a transmits the persona information to the terminal device 20 via the network 40.

  The persona information is received by the receiving unit 22b of the terminal device 20 and sent to the writing unit 23c (step S14). The writing unit 23c maintains the correspondence between each persona identifier (PID), each persona cluster identifier (CIDsel), and each second distribution information identifier (DID2sub) included in the sent persona information. Is stored in the distribution information identifier database storage unit 21b, and the distribution information identifier database is updated (step S15). FIG. 6B shows an example of the data structure of the distribution information identifier database updated in this way. As can be seen by comparing FIGS. 6A and 6B, the persona identifier and the persona cluster identifier are associated with each second distribution information identifier by this processing.

  As described above, the terminal device 20 can use the persona identifier corresponding to the persona cluster in the field suitable for each aspect of the user even when the user does not sufficiently act as the persona. The ID management technique using the persona cluster identifier can receive a service in consideration of the multiplicity of users.

[Second Embodiment]
Next, a second embodiment of the present invention will be described. This embodiment is a modification of the first embodiment, and further considers the similarity between users so that the information recommendation device recommends a persona. Below, it demonstrates centering around difference with 1st Embodiment, and abbreviate | omits description about the matter which is common in 1st Embodiment.

<Configuration>
The information recommendation system of the second embodiment is obtained by replacing the information recommendation device 10 of the information recommendation system of the first embodiment shown in FIG. Hereinafter, only the configuration of the information recommendation device 110 will be described.

  FIG. 9 is a block diagram illustrating a functional configuration of the information recommendation device 110 according to the second embodiment. FIG. 10 is a block diagram illustrating an example of a functional configuration of the user similarity determination unit 113d. In FIG. 9, the same reference numerals as those in FIG. 2 are used for the functional units common to the information recommendation device 10 of the first embodiment. Below, it demonstrates centering on difference with the information recommendation apparatus 10. FIG.

  As shown in FIG. 9, the information recommendation device 110 according to the present embodiment includes a persona cluster database storage unit 11a, a persona database storage unit 11b, a temporary storage unit 11c, a transmission unit 12a, a reception unit 12b, a control unit 13a, and information similarity determination. A unit 13b, a persona identifier assigning unit 13c, a user similarity determining unit 113d, and a persona cluster identifier adding unit 113e. As illustrated in FIG. 10, the user similarity determination unit 113d includes a user vector generation unit 113da, a correlation coefficient calculation unit 113db, and a selection unit 113dc. An example of the user similarity determination unit 113d and the persona cluster identifier addition unit 113e is a CPU in which a predetermined program is read.

<Processing>
Next, the processing of this embodiment will be described. Below, it demonstrates centering around difference with 1st Embodiment.
[Preprocessing]
Since it is the same as 1st Embodiment, description is abbreviate | omitted.
[Persona recommendation processing]
FIG. 11 is a flowchart for explaining the persona recommendation processing according to the second embodiment. Hereinafter, the persona recommendation processing of this embodiment will be described with reference to these drawings.

  First, the processes of steps S1 to S3 described in the first embodiment are executed. Each persona cluster identifier (CID) selected by the information similarity determination unit 13b in step S3 is sent to the user similarity determination unit 113d. For each user identifier (UID), the user similarity determination unit 113d includes a set of persona cluster identifiers (CID) associated with the same user identifier (UID) in the persona database of the persona database storage unit 11b. The similarity with the set of persona cluster identifiers (CIDs) extracted by the information similarity determination unit 13b is obtained, and one or more user identifiers (UIDs) are extracted in descending order of the similarity (step S21). . This is equivalent to evaluating the similarity between users.

<< Specific Example of Step S21 >>
In this example, first, each persona cluster identifier (CID) selected by the information similarity determination unit 13b in step S3 is input to the user vector generation unit 113da of the user similarity determination unit 113d. The user vector generation unit 113da generates a user vector (UV1) corresponding to the user of the terminal device 20 that recommends the persona using each persona cluster identifier (CID), and stores the user vector (UV1) in the temporary storage unit 11c. . The user vector (UV1) may be a vector indicating the characteristics of the set of persona cluster identifiers (CIDs) selected by the information similarity determination unit 13b in step S3. An example of the user vector (UV1) is a vector whose element is a value indicating whether or not each persona cluster identifier has been selected in step S3. For example, the following vector may be used as the user vector (UV1).
UV1 = {e ₁ , e ₂ , ..., e _K }… (1)

However, e _k = {0,1} (k = 1,..., K) is a value indicating whether or not the k-th persona cluster identifier (CID) existing in the persona cluster database is selected in step S3. It is. For example, if the kth persona cluster identifier (CID) present in the persona cluster database is selected in step S3, e _k = 1 is set, and if not selected, e _k = 0 is set. As another example, when the weight g _k is set for each persona cluster identifier (CID) and the k-th persona cluster identifier (CID) existing in the persona cluster database is selected in step S3, e _k = g _k may be set, and e _k = 0 when not selected. K is the total number of persona cluster identifiers (CIDs) present in the persona cluster database.

The user vector generation unit 113da of the user similarity determination unit 113d further includes, for each user identifier (UID) in the persona database, each persona cluster identifier (CID) corresponding to each user identifier (UID). ) Is input. The user vector generation unit 113da uses the set of persona cluster identifiers (CIDs), and for each user identifier (UID) of the persona database, each user vector (UV2) corresponding to each user identifier (UID). Are stored in the temporary storage unit 11c in association with the corresponding user identifier (UID). The user vector (UV2) is a vector indicating the characteristics of a set of persona cluster identifiers (CIDs) associated with the same user identifier (UID) in the persona database. However, the user vector (UV1) and the user vector (UV2) must indicate the characteristics of the set of persona cluster identifiers (CID) on the same basis. For example, if equation (1) is the user vector (UV1), the following vector needs to be the user vector (UV2).
UV2 = {f ₁ , f ₂ , ..., f _K }… (2)

However, f _k = {0,1} (k = 1, ..., K) means that the kth persona cluster identifier (CID) existing in the persona cluster database is the persona database, and the target user identifier. It is a value indicating whether or not it is associated with (UID). For example, if the k-th persona cluster identifier (CID) existing in the persona cluster database is associated with the target user identifier (UID) in the persona database, the association is made with f _k = 1. If not, set f _k = 0. As another example, when the weight g _k is set for each persona cluster identifier (CID) and the k-th persona cluster identifier (CID) existing in the persona cluster database is selected in step S3, f _k = g _k may be set, and f _k = 0 when not selected.

Next, the correlation coefficient calculation unit (not shown) of the user similarity determination unit 113d reads the user vector (UV1, 2) from the temporary storage unit 11c, and for each user identifier (UID), the user A correlation coefficient (COR) (corresponding to “similarity”) between the vector (UV1) and the user vector (UV2) is obtained. For example, if the user vectors (UV1, UV2) of equations (1) and (2) are used, the correlation coefficient (COR) is as follows.
COR = e ₁・ f ₁ + ... + e _K・ f _K … (3)
The generated correlation coefficient (COR) is stored in the temporary storage unit 11c in association with the corresponding user vector (UV2).

  Next, the selection unit 113dc compares each correlation coefficient (COR) stored in the temporary storage unit 11c, extracts one or more user identifiers (UID) in descending order of the correlation coefficient (COR), and outputs them. (<< End of description of specific example of step S21 >>).

  Next, the user identifier (UID) extracted by the user similarity determination unit 113d and each persona cluster identifier (CID) extracted by the information similarity determination unit 13b are input to the persona cluster identifier addition unit 113e. Is done. The persona cluster identifier adding unit 113e refers to the persona database, and each persona cluster identifier (CIDadd1) associated with the user identifier (UID) extracted by the user similarity determination unit 113d in the persona database. Each persona cluster identifier (CIDadd2) excluding each persona cluster identifier (CID) extracted by the information similarity determination unit 13b is extracted (step S22). Each extracted persona cluster identifier (CIDadd2) is sent to the transmission unit 12a. Further, each persona cluster identifier (CID) extracted by the information similarity determination unit 13b (step S3) and the corresponding second distribution information identifier (DID2sub) are input to the transmission unit 12a. The transmission part 12a transmits these to the terminal device 20 via the network 40 with each persona cluster identifier (CIDadd2) extracted by step S22 (step S24). They are received by the receiving unit 22a of the terminal device 20 (step S25). Thereafter, the processes of steps S6 to S15 described in the first embodiment are executed.

  FIG. 12A is a diagram illustrating a data configuration of a distribution information identifier database stored in the distribution information identifier database storage unit 21b of the terminal device 20 before performing the processing of the second embodiment. Moreover, FIG.12 (b) is the figure which illustrated the data structure of the delivery information identifier database stored in the delivery information identifier database memory | storage part 21b of the terminal device 20 after performing the process of 2nd Embodiment. In the case of this embodiment, not only the persona identifier is set based on the second distribution information identifier, but also another persona identifier is set based only on the similarity of the user. Therefore, in the data configuration example of FIG. 12B, there is also a persona identifier that is not associated with the second distribution information identifier.

  As described above, in this embodiment, the information recommendation device 110 further sets the persona identifier corresponding to the persona cluster in the field suitable for each aspect of the user based on the similarity of the user. In this case, even if the user is not acting as a persona as well as the individual granularity of the user, the user must identify the persona identifier corresponding to the persona cluster in the field that fits each aspect. Can be used.

[Third Embodiment]
Next, a third embodiment of the present invention will be described. This embodiment is a further modification of the second embodiment, in which the information recommendation device extracts a persona cluster identifier in consideration of the similarity between users, and further selects the extracted persona cluster identifier. This is performed using the distribution information identifiers of the users that were not used for the evaluation of the similarity between the two. Below, it demonstrates centering around difference with 1st, 2nd embodiment, and abbreviate | omits description about the matter which is common in 1st, 2nd embodiment.

<Configuration>
The information recommendation system of the third embodiment is obtained by replacing the information recommendation device 10 of the information recommendation system of the first embodiment shown in FIG. Below, only the structure of the information recommendation apparatus 210 is demonstrated.

  FIG. 13 is a block diagram illustrating a functional configuration of the information recommendation device 210 according to the third embodiment. FIG. 14 is a block diagram illustrating an example of a functional configuration of the selection unit 213b. In FIG. 13, the same reference numerals as those in FIGS. 2 and 9 are used for functional units common to the information recommendation devices 10 and 110 of the first and second embodiments. Below, it demonstrates centering around difference with the information recommendation apparatuses 10 and 110. FIG.

  As shown in FIG. 13, the information recommendation device 210 according to this embodiment includes a persona cluster database storage unit 11a, a persona database storage unit 11b, a temporary storage unit 11c, a transmission unit 12a, a reception unit 12b, a control unit 13a, and an information similarity determination. A unit 13b, a persona identifier assignment unit 13c, a user similarity determination unit 113d, a persona cluster identifier addition unit 113e, a difference set generation unit 213a, and a selection unit 213b. As illustrated in FIG. 14, the selection unit 213b includes a count unit 213ba and a comparison unit 213bb. An example of the difference set generation unit 213a and the selection unit 213b is a CPU in which a predetermined program is read.

<Processing>
Next, the processing of this embodiment will be described. Below, it demonstrates centering on difference with 1st, 2nd embodiment.
[Preprocessing]
Since it is the same as 1st Embodiment, description is abbreviate | omitted.
[Persona recommendation processing]
FIG. 15 is a flowchart for explaining a persona recommendation process according to the third embodiment. Hereinafter, the persona recommendation processing of this embodiment will be described with reference to these drawings.
First, the processes of steps S1 to S3 described in the first embodiment are executed. Next, the processes of steps S21 and S22 described in the second embodiment are executed. Next, the first distribution information respectively corresponding to the set of second distribution information identifiers (DID2) and each persona cluster identifier (CID) extracted by the information similarity determination unit 13b from the information similarity determination unit 13b. The set of the second distribution information identifier (DID2sub) that matches the identifier (DID1) is sent to the difference set generation unit 213a of the information recommendation device 210. The difference set generation unit 213a matches the first distribution information identifier (DID1) corresponding to each persona cluster identifier extracted by the information similarity determination unit from the set of second distribution information identifiers (DID2). A set of third distribution information identifiers (DID3) from which (DID2sub) has been removed is generated (step S31).

  The generated set of third distribution information identifiers (DID3) is sent to the selection unit 213b. Further, each persona cluster identifier (CIDadd2) extracted by the persona cluster identifier adding unit 113e is input to the selection unit 213b. The selection unit 213b refers to the persona cluster database, and for each persona cluster identifier, a set of a first distribution information identifier (DID1) corresponding to each persona cluster identifier (CIDadd2) and a third distribution information identifier (DID3) ) And a plurality of persona cluster identifiers (CIDadd3) are selected from the persona cluster identifiers (CIDadd2) extracted by the persona cluster identifier adding unit 113e and output in descending order of similarity ( Step S32).

<< Specific Example of Step S32 >>
In this example, first, the set of the third distribution information identifier (DID3) and the set of the individual persona cluster identifiers (CIDadd2) are input to the count unit 213ba of the selection unit 213b. The count unit 213ba refers to the persona cluster database, and for each persona cluster identifier (CIDadd2), a set of first distribution information identifiers (DID1) corresponding to each persona cluster identifier (CIDadd2) and third distribution information The set of identifiers (DID3) is compared, and the number of coincidence between the second distribution information identifier (DID1) and the third distribution information identifier (DID3) is counted. The count unit 213ba sets the count number as the similarity between the set of the first distribution information identifier (DID1) corresponding to each persona cluster identifier (CIDadd2) and the set of the third distribution information identifier (DID3). The count number (CT2) and the corresponding persona cluster identifier (CIDadd2) are stored in the temporary storage unit 11c in association with each other. The count unit 213ba performs a similar process for each persona cluster identifier (CIDadd2).

  The comparison unit 213bb compares the count numbers (CT2) respectively associated with the persona cluster identifier (CIDadd2) stored in the temporary storage unit 11c, and the count number (CT2) is large (similarity is high). In order, one or more persona cluster identifiers are selected as CIDadd3, and each selected persona cluster identifier (CIDadd3) is output.

  Note that the number of coincidence between the first distribution information identifier (DID1) corresponding to the persona cluster identifier (CIDadd2) and the third distribution information identifier (DID3) is simply counted, and the counted number is used as the similarity. Instead, a weight may be set for each first distribution information identifier (DID1), and a count number considering the weight may be used as the similarity. The number of persona cluster identifiers (CIDadd3) extracted by the comparison unit 213bb may be set in advance, or a threshold value regarding the count number (CT2) is set, and the count number (CT2) equal to or greater than the threshold value is set. All the persona cluster identifiers (CIDadd2) that have become may be extracted as CIDadd3 (<< end of description of specific example of step S32 >>).

  Next, the persona cluster identifier (CIDadd3) extracted in step S32 is sent to the transmission unit 12a. Further, each persona cluster identifier (CID) extracted by the information similarity determination unit 13b (step S3) and the corresponding second distribution information identifier (DID2sub) are input to the transmission unit 12a. The transmission part 12a transmits these to the terminal device 20 via the network 40 with each persona cluster identifier (CIDadd2) extracted by step S32 (step S34). They are received by the receiving unit 22a of the terminal device 20 (step S35). Thereafter, the processes of steps S6 to S15 described in the first embodiment are executed.

  As described above, in this embodiment, each persona cluster identifier extracted by the persona cluster identifier adding unit 113e transmitted from the transmission unit 12a of the information recommendation device 210 to the terminal device 20 is limited to the one selected by the selection unit 213b. It was decided to. As described above, for this selection, among the second distribution information identifiers (DID2) sent from the terminal device 20, the third distribution information identifier (DID3) that was not used for evaluating the similarity between users. ) Is used. As a result, persona cluster identifiers that more closely match the user's aspect can be selected and transmitted to the terminal device 20.

[Fourth Embodiment]
Next, a fourth embodiment of the present invention will be described. This embodiment is a modification of the second embodiment, and the information recommendation device extracts a user similar to the user based on the persona cluster identifier sent from the terminal device, and corresponds to the extracted user. The persona cluster identifier to be introduced is introduced. That is, this form introduces a new persona cluster identifier to a terminal device that already possesses at least one persona cluster identifier. Below, it demonstrates centering around difference with 1st, 2nd embodiment, and abbreviate | omits description about the matter which is common in 1st, 2nd embodiment.

<Configuration>
In the information recommendation system of the fourth embodiment, the information recommendation device 10 of the information recommendation system of the first embodiment shown in FIG. 1 is replaced with an information recommendation device 310, and the terminal device 20 is replaced with a terminal device 320. It is. Hereinafter, these differences will be mainly described.
FIG. 16 is a block diagram illustrating a functional configuration of the information recommendation device 310 according to the fourth embodiment. FIG. 17 is a block diagram illustrating a functional configuration of the terminal device 320 according to the fourth embodiment.

  As shown in FIG. 16, the information recommendation apparatus 310 of this embodiment includes a persona cluster database storage unit 11a, a persona database storage unit 11b, a temporary storage unit 11c, a transmission unit 12a, a reception unit 12b, a control unit 13a, and a persona identifier assignment unit. 13c, a user similarity determination unit 313d, and a persona cluster identifier addition unit 113e. Also, as shown in FIG. 17, the difference between the terminal device 320 of this embodiment and the terminal device 20 of the first embodiment is the flow of information.

<Processing>
Next, the processing of this embodiment will be described.
[Preprocessing]
The difference from the first embodiment is that the distribution information identifier database storage unit 21b of the terminal device 320 stores a persona identifier (PID) and a persona cluster identifier (CID). FIG. 18A is a diagram illustrating a data configuration of a distribution information identifier database stored in the distribution information identifier database storage unit 21b. Since others are the same as those of the first embodiment, the description thereof is omitted.

[Persona recommendation processing]
FIG. 19 is a flowchart for explaining a persona recommendation process according to the fourth embodiment. Hereinafter, the persona recommendation processing of this embodiment will be described with reference to these drawings.
First, a set of persona identifiers (PIDs) that specify a set of persona cluster identifiers (CIDs) stored in the distribution information identifier database storage unit 21b of the terminal device 320 is read and sent to the transmission unit 22a. Here, the set of persona cluster identifiers (CIDs) includes one or more persona cluster identifiers (CIDs), and the set of persona identifiers (PIDs) includes one or more persona identifiers (PIDs) associated with the persona identifiers (PIDs). ) As an element. The transmission unit 22a transmits the set of persona identifiers (PID) to the information recommendation device 310 via the network 40 (step S41).

  A set of persona identifiers (PID) is received by the receiving unit 12b of the information recommendation device 310 and sent to the user similarity determination unit 313d (step S42). The user similarity determination unit 313d refers to the persona database, and sets, for each user identifier (UID), a set of persona cluster identifiers (CIDs) associated with the same user identifier (UID) in the persona database. The similarity with the set of persona cluster identifiers (CID) specified by the persona identifier (PID) received by the receiving unit 12b is obtained, and one or more user identifiers (UID) are extracted in descending order of the similarity. (Step S43). This is equivalent to evaluating the similarity between users. Note that specific examples of how to obtain the similarity and the number of user identifiers (UIDs) to be extracted are the same as those of the user similarity determination unit 113d of the second embodiment.

  Each user identifier (UID) extracted by the user similarity determination unit 313d is sent to the persona cluster identifier addition unit 113e. Furthermore, the persona cluster identifier adding unit 113e is also sent the persona identifier (PID) received by the receiving unit 12b. The persona cluster identifier adding unit 113e refers to the persona database, and includes each persona cluster identifier (CID) associated with the user identifier (UID) extracted by the user similarity determination unit 313d in the persona database. Then, each persona cluster identifier (CID) excluding each persona cluster identifier (CID) corresponding to the persona identifier (PID) received by the receiving unit 12b is extracted (step S44).

  Each persona cluster identifier (CID) extracted by the persona cluster identifier adding unit 113e is sent to the transmission unit 12a, and the transmission unit 12a transmits each persona cluster identifier (CID) to the terminal device 320 via the network 40 (step). S45). The receiving unit 22a of the terminal device 320 receives each persona cluster identifier (CID), and then the processes of steps S6 to S15 described in the first embodiment are performed.

  FIG. 18B is a diagram illustrating a data configuration of a distribution information identifier database stored in the distribution information identifier database storage unit 21b of the terminal device 320 after performing the processing of the fourth embodiment. In the case of this embodiment, the persona identifier is set based only on the similarity of the user without being based on the second distribution information identifier. Therefore, in the data configuration example of FIG. 18B, a persona identifier that is not associated with the second distribution information identifier is newly added.

  As described above, in this embodiment, the information recommendation device 310 sets the persona identifier corresponding to the persona cluster in the field that is suitable for each aspect of the user based on the similarity of the user. In this case, the user possesses a certain amount of persona identifiers, but even if the persona is not acting sufficiently as a persona other than the corresponding persona, the user is not acting sufficiently. It becomes possible to newly use a persona identifier corresponding to a persona cluster in a field that conforms to.

[Fifth Embodiment]
Next, a fifth embodiment of the present invention will be described. In each of the above embodiments, first, the information recommendation device selects a persona cluster identifier based on the information sent from the terminal device, transmits them to the terminal device, and issues a persona cluster identifier for issuing a persona at the terminal device. After selecting, the persona corresponding to the selected persona cluster identifier was issued. In the fifth embodiment, the information recommendation device selects a persona cluster identifier based on information sent from the terminal device, and issues personas corresponding to all the selected persona cluster identifiers. The procedure for the information recommendation device to select a persona cluster identifier based on the information sent from the terminal device and the procedure for the information recommendation device to issue a persona corresponding to the persona cluster identifier will be described in any of the above embodiments. The procedure described above may be used.

[Sixth Embodiment]
Next, a sixth embodiment of the present invention will be described. This embodiment is a modification of the fifth embodiment. In the sixth embodiment, the information recommendation device selects a persona cluster identifier based on information transmitted from the terminal device, and issues personas corresponding to all the selected persona cluster identifiers. Further, the information recommendation device performs collaborative filtering for each extracted persona cluster identifier, and extracts a distribution information identifier recommended for the terminal device. Such a form is realizable as a modification of each of the first to fourth embodiments. However, a modification of the first embodiment will be described below as an example. In the following, the description will focus on differences from the above.

<Configuration>
The information recommendation system of the sixth embodiment is obtained by replacing the information recommendation device 10 of the information recommendation system of the first embodiment shown in FIG. Hereinafter, these differences will be mainly described.
FIG. 20 is a block diagram illustrating a functional configuration of the information recommendation apparatus 510 according to the sixth embodiment.
As shown in FIG. 20, the information recommendation device 510 of this embodiment includes a persona cluster database storage unit 11a, a persona database storage unit 11b, a temporary storage unit 11c, a transmission unit 12a, a reception unit 12b, a control unit 13a, and information similarity determination. A section 13b, a persona identifier assigning section 13c, and a recommended distribution information selecting section 513a. An example of the recommended distribution information selection unit 513a is a CPU in which a predetermined program is read.

<Processing>
Next, the processing of this embodiment will be described.
[Preprocessing]
Since it is the same as 1st Embodiment, description is abbreviate | omitted.
[Persona recommendation processing]
FIG. 21 is a flowchart for explaining a persona recommendation process according to the sixth embodiment. Hereinafter, the persona recommendation processing of this embodiment will be described with reference to this figure.
First, the processes of steps S1 to S3 described in the first embodiment are executed. The persona cluster identifier (CID) extracted by the information similarity determination unit 13b in step S3 and the second distribution information identifier (DID2) received by the reception unit 12b in step S2 are sent to the recommended distribution information selection unit 513a. . For each persona cluster identifier (CID) extracted by the information similarity determination unit 13b, the recommended distribution information selection unit 513a refers to the persona cluster database, and is associated with the persona cluster identifier (CID) in the persona cluster database. A fourth distribution information identifier (DID4) that is each first distribution information identifier (DID1) and excludes each second distribution information identifier (DID2) is extracted (step S61). The recommended distribution information selection unit 513a sends the extracted fourth distribution information identifiers (DID4) and the corresponding persona cluster identifiers (CIDs) to the transmission unit 12a in association with each other. In addition, the transmission unit 12a includes a second person ID that matches the persona cluster identifier (CID) extracted in step S3 and the first distribution information identifier (DID1) associated with the persona cluster identifier (CID). The distribution information identifier (DID2sub) is sent in association with each other. The transmission part 12a transmits each transmitted information to said terminal device 20 via the network 40 (step S64).

  Each of these pieces of information is received by the receiving unit 22b of the terminal device 20 (step S65). The received fourth distribution information identifiers (DID4), the corresponding persona cluster identifiers (CID) and the second distribution information identifiers (DID2sub) are sent to the writing unit 23c, and the writing unit 23c Stores them in the distribution information identifier database storage unit 21b in association with each other, and updates the distribution information identifier database (step S66). Thereafter, the processes of steps S6 to S15 described in the first embodiment are executed.

  As described above, the information recommendation device 510 according to the present embodiment uses the persona cluster identifier (CID) extracted for the terminal device 20 to perform collaborative filtering in units of persona cluster identifier (CID) (step S61), and the terminal The fourth distribution information identifier (DID4) is recommended to the device 20. As a result, collaborative filtering with persona granularity is possible, and distribution information that matches each aspect of the user can be provided.

[Seventh Embodiment]
Next, a seventh embodiment of the present invention will be described. This embodiment is a modification of the first to sixth embodiments and recommends distribution information to a terminal device based on a persona cluster identifier already set by the information recommendation device. The processing of this embodiment is based on the premise that the distribution information identifier database storage unit 21b of the user device 20 stores a persona identifier corresponding to the second distribution information identifier and a persona cluster identifier in association with each other. . For example, in the process of this embodiment, the persona identifier corresponding to the second distribution information identifier and the persona cluster identifier correspond to the distribution information identifier database storage unit 21b of the user device 20 by the process of any of the above embodiments. It is executed after being attached and stored. Below, it demonstrates focusing on difference with said each embodiment.

<Configuration>
The information recommendation system of the seventh embodiment is obtained by replacing the information recommendation device 10 of the information recommendation system of the first embodiment shown in FIG. Only the configuration of the information recommendation device 610 will be described below.

  FIG. 22 is a block diagram illustrating a functional configuration of the information recommendation device 610 according to the seventh embodiment. Note that FIG. 22 shows only functional units related to processing for recommending distribution information to the terminal device 20 based on the persona cluster identifier already set by the information recommendation device 610. Since the other functional units are as described above, description thereof is omitted.

  As shown in FIG. 22, the information recommendation device 610 of this embodiment includes a persona cluster database storage unit 11a, a persona database storage unit 11b, a temporary storage unit 11c, a transmission unit 12a, a reception unit 12b, a control unit 13a, and recommended distribution information. A selection unit 613a is included. An example of the difference recommended distribution information selection unit 613a is a CPU in which a predetermined program is read.

<Processing>
FIG. 23 is a flowchart for explaining the processing of the seventh embodiment. The processing of this embodiment will be described using this figure.
First, a persona identifier (PID) stored in the distribution information identifier database storage unit 21b of the terminal device 20 and a set of fifth distribution information identifiers (DID5) associated therewith are read and transmitted. Sent to the unit 22a. The persona identifier (PID) read out here may be all persona identifiers stored in the distribution information identifier database storage unit 21b or a part of the persona identifiers. The transmission unit 22a transmits the persona identifier (PID) and a set of fifth distribution information identifiers (DID5) associated therewith to the information recommendation device 610 via the network 40 (step S71).

  The persona identifier (PID) and the set of fifth distribution information identifiers (DID5) associated therewith are received by the receiving unit 12b of the information recommendation device 610 and sent to the recommended distribution information selecting unit 613a (step S72). ). The recommended distribution information selection unit 613a first refers to the persona database and identifies a persona cluster identifier (CID) associated with the persona identifier (PID) received by the receiving unit 12b (step S73). Next, the recommended distribution information selection unit 613a refers to the persona cluster database, and each first distribution information identifier (DID1) associated with the persona cluster identifier (CID) specified as described above in the persona cluster database. Then, the fourth distribution information identifier (DID4) excluding each fifth distribution information identifier (DID5) is extracted (step S74). The extracted fourth distribution information identifier (DID4) is sent to the transmission unit 12a in association with the corresponding persona identifier (PID). The transmitter 12a transmits these pieces of information to the terminal device 20 via the network 40 (step S75).

  Each of these pieces of information is received by the receiving unit 22b of the terminal device 20 (step S76). Each received fourth distribution information identifier (DID4) and each corresponding persona identifier (PID) are sent to the writing unit 23c, and the writing unit 23c associates them with the distribution information identifier database. The distribution information identifier database is updated in the storage unit 21b (step S77).

  As described above, the information recommendation device 510 according to the present embodiment uses the persona cluster identifier (CID) extracted for the terminal device 20 to perform collaborative filtering in units of persona cluster identifier (CID) (step S74), and the terminal The fourth distribution information identifier (DID4) is recommended to the device 20. As a result, collaborative filtering with persona granularity is possible, and distribution information that matches each aspect of the user can be provided.

  In addition, this invention is not limited to the above-mentioned embodiment. For example, the various processes described above are not only executed in time series according to the description, but may also be executed in parallel or individually as required by the processing capability of the apparatus that executes the processes. Needless to say, other modifications are possible without departing from the spirit of the present invention.

Further, when the above-described configuration is realized by a computer, processing contents of functions that each device should have are described by a program. The processing functions are realized on the computer by executing the program on the computer.
The program describing the processing contents can be recorded on a computer-readable recording medium. As the computer-readable recording medium, for example, any recording medium such as a magnetic recording device, an optical disk, a magneto-optical recording medium, and a semiconductor memory may be used.

  The program is distributed by selling, transferring, or lending a portable recording medium such as a DVD or CD-ROM in which the program is recorded. Furthermore, the program may be distributed by storing the program in a storage device of the server computer and transferring the program from the server computer to another computer via a network.

  A computer that executes such a program first stores, for example, a program recorded on a portable recording medium or a program transferred from a server computer in its storage device. When executing the process, the computer reads a program stored in its own recording medium and executes a process according to the read program. As another execution form of the program, the computer may directly read the program from a portable recording medium and execute processing according to the program, and the program is transferred from the server computer to the computer. Each time, the processing according to the received program may be executed sequentially. Also, the program is not transferred from the server computer to the computer, and the above-described processing is executed by a so-called ASP (Application Service Provider) type service that realizes the processing function only by the execution instruction and result acquisition. It is good. Note that the program in this embodiment includes information that is used for processing by an electronic computer and that conforms to the program (data that is not a direct command to the computer but has a property that defines the processing of the computer).

  In this embodiment, the present apparatus is configured by executing a predetermined program on a computer. However, at least a part of these processing contents may be realized by hardware.

  As an industrial application field of the present invention, for example, a persona setting field for a product information introduction service in persona units on the Internet can be exemplified.

FIG. 1 is a conceptual diagram illustrating the configuration of the information recommendation system according to the first embodiment. FIG. 2 is a block diagram illustrating a functional configuration of the information recommendation device according to the first embodiment. FIG. 3 is a block diagram illustrating an example of a functional configuration of the information similarity determination unit according to the first embodiment. FIG. 4 is a block diagram illustrating a functional configuration of the terminal device according to the first embodiment. FIG. 5A is a diagram illustrating a data configuration of a persona cluster database stored in the persona cluster database storage unit. FIG. 5B is a diagram illustrating a data configuration of the persona database stored in the persona database storage unit. FIG. 6A is a diagram illustrating a data configuration of a distribution information identifier database stored in a user identifier storage unit of a terminal device to which no persona identifier is assigned. FIG. 6B shows an example of the data structure of the distribution information identifier database updated in this way. FIG. 7 is a flowchart for explaining the persona recommendation process according to the first embodiment. FIG. 8 is a flowchart for explaining the persona recommendation process according to the first embodiment. FIG. 9 is a block diagram illustrating a functional configuration of the information recommendation device according to the second embodiment. FIG. 10 is a block diagram illustrating an example of a functional configuration of the user similarity determination unit. FIG. 11 is a flowchart for explaining the persona recommendation processing according to the second embodiment. FIG. 12A is a diagram illustrating a data configuration of a distribution information identifier database stored in the distribution information identifier database storage unit of the terminal device before performing the processing of the second embodiment. FIG. 12B is a diagram illustrating a data configuration of a distribution information identifier database stored in the distribution information identifier database storage unit 21b of the terminal device after the processing of the second embodiment is performed. FIG. 13 is a block diagram illustrating a functional configuration of the information recommendation device according to the third embodiment. FIG. 14 is a block diagram illustrating an example of a functional configuration of the selection unit. FIG. 15 is a flowchart for explaining a persona recommendation process according to the third embodiment. FIG. 16 is a block diagram illustrating a functional configuration of the information recommendation device according to the fourth embodiment. FIG. 17 is a block diagram illustrating a functional configuration of a terminal device according to the fourth embodiment. FIG. 18A is a diagram illustrating a data configuration of a distribution information identifier database stored in the distribution information identifier database storage unit 21b. FIG. 18B is a diagram illustrating a data configuration of the distribution information identifier database stored in the distribution information identifier database storage unit of the terminal device after performing the processing of the fourth embodiment. FIG. 19 is a flowchart for explaining a persona recommendation process according to the fourth embodiment. FIG. 20 is a block diagram illustrating a functional configuration of the information recommendation device according to the sixth embodiment. FIG. 21 is a flowchart for explaining a persona recommendation process according to the sixth embodiment. FIG. 22 is a block diagram illustrating a functional configuration of the information recommendation device according to the seventh embodiment. FIG. 23 is a flowchart for explaining the processing of the seventh embodiment.

Explanation of symbols

1 Information recommendation system 10, 110, 210, 310, 510, 610 Information recommendation device 20, 320 Terminal device

Claims (10)

Information recommendation to send information to terminal devices to which information is distributed through the network, in order to recommend a persona that is a different personality for each field of the distribution information when the user receives information distribution through the network A device,
Each persona cluster identifier that identifies a plurality of clusters obtained by classifying a plurality of pieces of distribution information that can be distributed through the network, and a first distribution that identifies each piece of distribution information that belongs to a cluster corresponding to the persona cluster identifier. A first storage unit for storing a persona cluster database in which an information identifier is associated;
A receiving unit that receives from the terminal device a set of second distribution information identifiers that respectively identify a plurality of pieces of distribution information;
For each persona cluster identifier, obtain a similarity between the first set of distribution information identifiers associated with the same persona cluster identifier in the persona cluster database and the second set of distribution information identifiers, An information similarity determination unit that selects one or more persona cluster identifiers in descending order of the similarity, and extracts the selected persona cluster identifier;
A transmission unit that transmits the persona cluster identifiers extracted by the information similarity determination unit to the terminal device;
An information recommendation device. The information recommendation device according to claim 1,
The information recommendation device further includes:
Stores a persona database in which a user identifier for identifying each user is associated with a persona cluster identifier for identifying the cluster to which the distribution information identifier for identifying distribution information distributed to the terminal device of the user belongs. A second storage unit to
For each user identifier, a set of each persona cluster identifier associated with the same user identifier in the persona database, and a set of each persona cluster identifier extracted by the information similarity determination unit A user similarity determination unit for obtaining a similarity and extracting one or more user identifiers in descending order of the similarity;
Each persona cluster identifier associated in the persona database with the user identifier extracted by the user similarity determination unit, and each persona cluster identifier extracted by the information similarity determination unit is A persona cluster identifier adding unit for extracting each persona cluster identifier excluding,
The transmitter further includes:
Transmitting each persona cluster identifier extracted by the persona cluster identifier adding unit to the terminal device;
An information recommendation device characterized by that. The information recommendation device according to claim 2, further comprising:
Third distribution information obtained by removing a set of the second distribution information identifiers that matches the first distribution information identifier corresponding to each persona cluster identifier extracted by the information similarity determination unit. A difference information generator for generating a set of identifiers;
For each persona cluster identifier, a similarity between the first distribution information identifier set corresponding to each persona cluster identifier extracted by the persona cluster identifier addition unit and the third distribution information identifier set is obtained. A selection unit that selects a plurality of persona cluster identifiers from each persona cluster identifier extracted by the persona cluster identifier addition unit in descending order of similarity;
Each persona cluster identifier extracted by the persona cluster identifier adding unit that is transmitted to the terminal device by the transmission unit is selected by the selection unit.
An information recommendation device characterized by that. Information recommendation to send information to terminal devices to which information is distributed through the network, in order to recommend a persona that is a different personality for each field of the distribution information when the user receives information distribution through the network A device,
A persona database in which a user identifier for identifying each user and a persona cluster identifier for identifying a cluster to which a distribution information identifier for identifying distribution information distributed to the terminal device of the user belongs is stored. A storage unit;
A receiving unit for receiving information identifying a set of persona cluster identifiers transmitted from the terminal device;
For each user identifier, a set of persona cluster identifiers associated with the same user identifier in the persona database, and a set of persona cluster identifiers specified by information received by the receiving unit; A user similarity determination unit that calculates one or more user identifiers in descending order of the similarity,
Each persona cluster identifier associated with the user identifier extracted by the user similarity determination unit in the persona database and specified by the information received by the receiving unit A persona cluster identifier adding unit for extracting each persona cluster identifier excluding the identifier;
A transmitting unit that transmits each persona cluster identifier extracted by the persona cluster identifier adding unit to the terminal device;
An information recommendation device. The information recommendation device according to any one of claims 1 to 4,
For each of the extracted persona cluster identifiers, a fourth distribution that is the first distribution information identifier associated with the persona cluster identifier in the persona cluster database, excluding the second distribution information identifier. A recommended delivery information selection unit for extracting an information identifier;
The transmission unit further transmits the fourth distribution information identifiers extracted for each persona cluster identifier by the recommended distribution information selection unit to the terminal device for each persona cluster identifier. ,
An information recommendation device characterized by that. The information recommendation device according to any one of claims 1 to 4,
The receiver further includes
Receiving information identifying a persona cluster identifier transmitted from the terminal device, and each fifth distribution information identifier corresponding to each distribution information,
The information recommendation device further includes:
Each of the first distribution information identifiers associated with the persona cluster identifier specified by the information received by the receiving unit in the persona cluster database, except for the fifth distribution information identifier. A recommended distribution information selection unit for extracting the distribution information identifier of
The transmission unit further transmits each fourth distribution information identifier extracted by the recommended distribution information selection unit to the terminal device.
An information recommendation device characterized by that. Information recommendation to send information to terminal devices to which information is distributed through the network, in order to recommend a persona that is a different personality for each field of the distribution information when the user receives information distribution through the network A device information recommendation method,
Each persona cluster identifier that identifies a plurality of clusters obtained by classifying a plurality of pieces of distribution information that can be distributed through the network, and a first distribution that identifies each piece of distribution information that belongs to a cluster corresponding to the persona cluster identifier. Store the persona cluster database that associates the information identifier with the storage unit,
Receiving a second set of distribution information identifiers for identifying a plurality of pieces of distribution information from the terminal device;
The information similarity determination unit includes, for each persona cluster identifier, the first set of distribution information identifiers associated with the same persona cluster identifier in the persona cluster database, and the second set of distribution information identifiers. Obtaining one or more persona cluster identifiers in descending order of the degree of similarity, and extracting the selected persona cluster identifiers;
A process of transmitting each persona cluster identifier extracted by the information similarity determination unit to the terminal device;
Information recommendation method to execute. Information recommendation to send information to terminal devices to which information is distributed through the network, in order to recommend a persona that is a different personality for each field of the distribution information when the user receives information distribution through the network A device information recommendation method,
A persona database in which a user identifier for identifying each user and a persona cluster identifier for identifying a cluster to which a distribution information identifier for identifying distribution information distributed to the terminal device of the user belongs are associated with each other. Stored in the
A receiving unit receiving a set of persona cluster identifiers transmitted from the terminal device;
A user similarity determination unit, for each user identifier, a set of persona cluster identifiers associated with the same user identifier in the persona database, and each persona cluster received by the receiver Obtaining a similarity with a set of identifiers and extracting one or more user identifiers in descending order of the similarity;
The persona cluster identifier adding unit is each persona cluster identifier associated with the user identifier extracted by the user similarity determination unit in the persona database, and each of the persona cluster identifiers received by the receiving unit Extracting each persona cluster identifier excluding the persona cluster identifier;
A process of transmitting each persona cluster identifier extracted by the persona cluster identifier adding unit to the terminal device;
Information recommendation method to execute.   The program for functioning a computer as an information recommendation apparatus in any one of Claim 1 to 6.   A computer-readable recording medium storing the program according to claim 9.

Images