JP2007233925A - Information processor and information processing method, program and recording medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To accurately analyze magnitude of influence of each node to another node in a network. <P>SOLUTION: A content intermediation centricity calculation part 61 calculates content intermediation centricity of each content on the basis of a content network 43, and supplies values thereof to an arithmetic part 63. A user intermediation centricity calculation part 62 performs processing similar to the case of the content intermediation centricity calculation part 61 to a user network 44, calculate user intermediation centricity of each user, and supplies values thereof to the arithmetic part 63. The arithmetic part 63 calculates a user importance level showing an importance level based of intensity of the influence of each the user to another user on the basis of the user intermediation centricity and the content intermediation centricity. This invention can be applied to an information processor. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an information processing apparatus and method, a program, and a recording medium, and in particular, an information processing apparatus capable of more accurately analyzing the magnitude of influence of each node on other nodes in a network. The present invention relates to a method, a program, and a recording medium.

  Conventionally, there is data mining that analyzes a large amount of data using various statistical analysis techniques to find hidden relationships and meanings for the purpose of sales promotion and information dissemination efficiency. For example, there is a market basket analysis in which POS data and EC site transaction data are analyzed to find a combination of “products purchased together”. This analysis result is used as a reference when considering, for example, in-store layout, shelving allocation, display planning, selection of special sale / commercial sale campaign products, and product purchase.

  In addition, as a recommendation technique used for recommending (advertisement), etc. of products and information, past user preference patterns are analyzed based on user history (for example, Internet user site browsing history, click history, etc.), Based on the analysis result, the current user's preference is estimated, and there are also recommendations using products and information recommended by the user and collaborative filtering. Collaborative filtering is a technique for recording a user's preference in the form of past behavior and inferring the user's preference based on the preference information of the user who is taking a behavior similar to that user.

  Such analysis is an important technique for determining marketing policies such as sales, sales, research, sales promotion activities, promotions, activities, and customer service.

  For example, as an extension of collaborative filtering, out of two users who are scheduled to record the same program, one user is scheduled to record simultaneously with that program, and the other user is not scheduled to record. There is an information processing apparatus that creates information that matches the user's preference, such as recommending it to other users, and provides information that matches the user's preference (see, for example, Patent Document 1). By performing sales promotion activities using the analysis results in this way, it is possible to expect effects such as increased access and increased sales.

  By the way, in recent years, with the improvement of information technology, information devices such as personal computers and portable terminal devices, information networks such as public line networks and the Internet have become widespread, and at the same time, such as WEB bulletin boards and SNS (Social Network Sites) CGM (Consumer Generated Media) is increasing, and information transmission by individuals and information exchange between individuals have been actively performed.

  As a result, for example, the so-called “word of mouth” consumer, customer, or user marketing target has become a node, and the information distribution network scale and amount of information performed via an information network such as the Internet has increased. The influence on the sale of products and the spread of information is also increasing.

Therefore, in recent years, marketing using such information distribution between individuals is also performed. For example, by conducting promotions and sales promotion activities such as providing detailed product information and trial products to users who actively communicate and exchange such information, the recognition of products and the number of purchases are increased. There is a way to expect.
JP-A-2005-57713

  However, in marketing using such information distribution between individuals, it is important to distribute information (for example, information on products) widely (many), but in general such information distribution Since the number of individuals involved is unspecified, there are various patterns of preference and behavior, and there is a possibility that the expected effect cannot be achieved even if the analysis method as described above is used.

  For example, when recommending a product B derived based on a method such as market basket analysis or collaborative filtering to a user who purchased product A, the recommendation is for a user whose preference pattern is similar to that user. Even if it is effective, it is not effective for other users whose preference patterns are not similar, and there is a possibility that information distribution is limited to some users.

  Further, for example, even if a user purchases many products, if the amount of information transmitted is small, the influence of the user on information distribution may be small. In addition, for example, even when information is actively transmitted on a web bulletin board with a large number of accesses, the accessing user is limited to a specific small number of people (particularly a small number of people with a specific preference). , The impact on the information distribution of the web bulletin board may be small.

  As described above, in the information distribution between individuals such as word-of-mouth communication, there is a fear that efficient marketing cannot be performed even if the analysis method as described above is used.

  The present invention has been made in view of such a situation, and by analyzing the magnitude of the influence of each node on other nodes in the network more accurately, for example, for information distribution between individuals. It is intended to enable effective marketing even for them.

  One aspect of the present invention is that both the first parameter and the second parameter are nodes, and only the first parameter is a node from a mixed network in which the relationship between the first parameter and the second parameter is shown as a link connecting the nodes. A first parameter network, a network generation unit that generates a second parameter network having only the second parameter as a node, and a first parameter network and a second parameter network generated by the network generation unit, The information processing apparatus includes a second parameter importance calculating unit that calculates the importance of the second parameter using the strength of influence as an index.

  The second parameter importance calculating means calculates a first parameter mediated centrality indicating the strength of influence of the first parameter on the other based on the first parameter network generated by the network generating means. Based on the parameter-mediated centrality calculating means and the second parameter network generated by the network generating means, the second parameter-mediated centrality for calculating the second parameter-mediated centrality indicating the strength of influence on the other of the second parameter. Based on the centrality calculating means, the first parameter-mediated centrality calculated by the first parameter-mediated centrality calculating means, and the second parameter-mediated centrality calculated by the second parameter-mediated centrality calculating means And calculating means for calculating the importance of the second parameter.

  In the first parameter network, the first parameter mediated centrality calculating means obtains a ratio of paths passing through the target node among the shortest paths between the other two nodes of the target node, and calculates the ratio to the other two nodes. The first parameter mediated centrality is calculated by averaging all the combinations of the first parameter mediated centrality, and the second parameter mediated centrality calculating means is a second parameter network in the shortest path between the other two nodes of the target node in the second parameter network. The second parameter mediated centrality is calculated by calculating a ratio of paths passing through the target node and averaging the ratio for all combinations of the other two nodes. Of the first parameter corresponding to the first parameter calculated by the first parameter-mediated centrality calculating means Calculating the sum of the through centrality, and the sum, the operation for calculating the sum of the second parameter mediated centrality calculated by the second parameter mediated centrality calculating means, can be carried out respectively, for each of the second parameter.

  The first parameter mediated centrality calculating means calculates the ratio of the first path through the target node among the shortest paths between the other two nodes of the target node in the first parameter network. The parameter can be multiplied by a weighting factor of an index different from the first parameter-mediated centrality.

  Storage means for storing first parameter-related information regarding the first parameter, second parameter-related information regarding the second parameter, and the first parameter network and the second parameter network, which are information for generating the mixed network Can further be provided.

  The apparatus may further comprise second parameter importance degree utilization information generation means for generating second parameter utilization information utilizing the importance degree of the second parameter calculated by the second parameter importance degree calculation means.

  The second parameter importance calculation means further calculates the importance of the first parameter using the strength of influence on others as an index based on the first parameter network generated by the network generation means, The information processing apparatus may further include first parameter importance degree utilization information generation means for generating first parameter utilization information utilizing the importance degree of the first parameter calculated by the parameter importance degree calculation means.

  One aspect of the present invention is that both the first parameter and the second parameter are nodes, and only the first parameter is a node from a mixed network in which the relationship between the first parameter and the second parameter is shown as a link connecting the nodes. A first parameter network and a second parameter network each having only the second parameter as a node are generated, and the second parameter network and the second parameter network are used to indicate the strength of influence on others as an index. This is an information processing method for executing a step of calculating the importance of a parameter.

  A first parameter network having only the first parameter as a node from a mixed network in which both the first parameter and the second parameter are nodes, and the relationship between the first parameter and the second parameter is shown as a link connecting the nodes; A second parameter network having only two parameters as nodes is generated, and the importance of the second parameter is calculated using the first parameter network and the second parameter network as indicators of the strength of influence on others. be able to.

  It can be set as the recording medium with which the program of Claim 9 is recorded.

  In one aspect of the present invention, both the first parameter and the second parameter are nodes, and only the first parameter is a node from a mixed network in which the relationship between the first parameter and the second parameter is shown as a link connecting the nodes. The first parameter network and the second parameter network having only the second parameter as nodes are respectively generated, and the generated first parameter network and the second parameter network are used to indicate the strength of influence on others. The importance of the second parameter is calculated.

  According to an aspect of the present invention, information can be analyzed. In particular, the magnitude of the influence of each node on other nodes in the network can be analyzed more accurately.

  According to an aspect of the present invention, for example, in a network in which a user distributes content information, the network structure is analyzed for content information distribution, the importance of each content regarding information distribution, and the user's content information distribution The importance regarding can be calculated. This makes it possible to know important content for distributing a large amount of content, to know important users for that purpose, and to provide information that supports effective analysis of users and content. Moreover, the analysis can be performed, and a more effective service based on the analysis result can be provided to the network.

  Embodiments of the present invention will be described below. Correspondences between the configuration requirements of the present invention and the embodiments described in the detailed description of the present invention are exemplified as follows. This description is to confirm that the embodiments supporting the present invention are described in the detailed description of the invention. Accordingly, although there are embodiments that are described in the detailed description of the invention but are not described here as embodiments corresponding to the constituent elements of the present invention, It does not mean that the embodiment does not correspond to the configuration requirements. Conversely, even if an embodiment is described here as corresponding to a configuration requirement, that means that the embodiment does not correspond to a configuration requirement other than the configuration requirement. It's not something to do.

  Further, this description does not mean all the inventions described in this specification. In other words, this description is an invention described in the present specification and is not claimed in this application, that is, an invention that will be filed in the future or added by amendment. Is not to deny.

  According to an aspect of the present invention, both the first parameter (for example, content C1 to content C5 in FIG. 3) and the second parameter (for example, user u1 to user u4 in FIG. 3) are nodes, and the first parameter and the second parameter A first parameter network (for example, the content network 43 in FIG. 4) having only the first parameter as a node from a mixed network (for example, the mixed network 45 in FIG. 3) whose relationship with the parameters is shown as a link connecting the nodes; Network generation means (for example, the network generation unit 31 in FIG. 1) for generating a second parameter network (for example, the user network 44 in FIG. 5) each having only the second parameter as a node, and the first generated by the network generation means. 1 parameter network and 2nd parameter network The second parameter importance calculating means (for example, the user in FIG. 1) that calculates the importance (for example, the user importance in FIG. 6) of the second parameter using the network as an index of the strength of influence on others. 1 is an information processing apparatus (for example, the analysis support apparatus 10 having the information processing section 11 in FIG. 1).

The second parameter importance calculation means is based on the first parameter network generated by the network generation means, and the first parameter mediated centrality (for example, the expression (1) 1) Ce (n _i )) for calculating first parameter-mediated centrality calculating means (for example, the content-mediated centrality calculating unit 61 in FIG. 6) and the second parameter network generated by the network generating means. A second parameter mediated centrality calculating unit (for example, a user mediated centrality calculating unit 62 in FIG. 6) for calculating a second parameter mediated centrality indicating the strength of influence on the other of the second parameter; The first parameter mediated centrality calculated by the parameter mediated centrality calculating means and the second parameter calculated by the second parameter mediated centrality calculating means. Over data mediated centrality (e.g., UC (u _i) of the formula (3)) based on, performs a predetermined calculation, importance of the second parameter (e.g., IC of formula (3) (u _i)) An arithmetic means for calculating (for example, the arithmetic part 63 in FIG. 6) can be provided.

In the first parameter network, the first parameter-mediated centrality calculating means is a ratio of paths passing through the target node among the shortest paths between the other two nodes of the target node (for example, g _{i in} Expression (1)). ^(st) / n _st ), and the ratio is averaged for all combinations of the other two nodes to calculate the first parameter-mediated centrality, and the second parameter-mediated centrality calculating means In the network, the ratio of the path passing through the target node in the shortest path between the other two nodes of the target node is obtained, and the ratio is averaged for all combinations of the other two nodes. The centrality is calculated, and the computing means calculates the first parameter-mediated centrality of the first parameter corresponding to the second parameter in the mixed network. First parameter mediated centrality of the sum calculated by means (e.g., CB (u _j) of the formula (2)) is calculated, the sum and the second parameter mediated calculated by the second parameter mediated centrality calculating means An operation for calculating the sum of centrality can be performed for each second parameter.

The first parameter mediated centrality calculating means calculates the ratio of the first path through the target node among the shortest paths between the other two nodes of the target node in the first parameter network. The parameter can be multiplied by a weighting factor of an index different from the first parameter-mediated centrality (for example, the weighting factor w _{st in} equation (21)).

  The first parameter related information related to the first parameter (for example, the content information 42 in FIG. 1) and the second parameter related information related to the second parameter (for example, the user in FIG. 1), which are information for generating the mixed network. The log 41) and storage means (for example, the database 14 of FIG. 1) for storing the first parameter network and the second parameter network may be further provided.

  Second parameter importance utilization information generating means for generating second parameter utilization information utilizing the importance of the second parameter calculated by the second parameter importance calculation means (for example, user importance utilization information generation in FIG. 1) Part 22) may further be provided.

  The second parameter importance calculation means further calculates the importance of the first parameter using the strength of influence on others as an index based on the first parameter network generated by the network generation means, First parameter importance utilization information generating means for generating first parameter utilization information utilizing the importance of the first parameter calculated by the parameter importance calculation means (for example, the content importance utilization information generating unit 122 in FIG. 11A). Can further be provided.

  One aspect of the present invention is that both the first parameter and the second parameter are nodes, and only the first parameter is a node from a mixed network in which the relationship between the first parameter and the second parameter is shown as a link connecting the nodes. A first parameter network and a second parameter network having only the second parameter as nodes are generated (for example, step S21 in FIG. 8), and the first parameter network and the second parameter network are used to influence the others. An information processing method (for example, an information processing method including the importance analysis processing method of FIG. 8) that executes the step of calculating the importance of the second parameter using the strength of the index (for example, step S22 of FIG. 8). is there.

  Next, an embodiment to which the present invention is applied will be described with reference to the drawings.

  FIG. 1 is a block diagram showing a configuration example of an analysis support apparatus to which the present invention is applied.

  In FIG. 1, an analysis support apparatus 10 is an information processing apparatus used by a user 1 who performs marketing such as sales, sales, research, sales promotion activities, promotion activities, and customer services.

  The analysis support apparatus 10 analyzes the importance of the first parameter as a target for distributing information and the second parameter as a medium for information distribution from the viewpoint of ease of distribution. At this time, for the first parameter, the analysis support apparatus 10 assumes that the higher the degree of influence on the other, the higher the importance, and the second parameter, the information of the higher importance, Make media that is more widely distributed more important. In the following description, the first parameter is assumed to be content, and the second parameter is assumed to be a user.

  The user 1 performs analysis based on the content calculated in this way by the analysis support apparatus 10 and information on the importance of the user, and determines a marketing policy and the like.

  For example, the analysis support apparatus 10 uses a content as a product, a user as a purchaser of the product, what is a product that can be sold to a wide range of customers, and who is a user who purchases more of such a product. The importance level of the purchaser is calculated and the information is provided to the user 1. With this information, the user 1 identifies, for example, a purchaser who is highly important for marketing, that is, a purchaser who has a strong influence on other purchasers, and regards the purchaser as the center of the purchaser group. Customer services such as discounts and preferential treatment can be developed so that the purchaser becomes the center of the service.

  Here, a buyer who has a strong influence on other purchasers is a purchaser whose marketing activities have a strong influence on the entire purchaser when the focus is on the target of marketing. Yes, it does not necessarily mean that you have a strong voice over other buyers. In general, a buyer who has a typical purchase pattern of a product group (a buyer who purchases more products purchased by more buyers) has a strong influence on other buyers. In many cases. The user 1 can improve the influence of the marketing activity on the entire purchaser by performing the marketing activity with such a purchaser as the center of the purchaser group (average purchase pattern). That is, the user 1 can perform marketing activities more efficiently by performing marketing activities based on the information obtained from the analysis support apparatus 10.

  In addition, for example, the analysis support apparatus 10 uses content as a product, sets a user as a WEB site on which reports on the product are posted, and what is a product that can be sold to a wide range of customers, and more reports on such products. From the perspective of where the web site is, the importance of the product or web site is calculated and the information is provided to the user 1. With this information, for example, the user 1 can specify a website that is the center of the website group in marketing, and can perform marketing activities such as sales promotion activities and advertising activities around the website. Thereby, the user 1 can perform marketing activities more efficiently, for example, by increasing the amount of information related to important products on the important website.

  Further, for example, the analysis support apparatus 10 uses a web bulletin board as a content, a user as a writing user who writes on the web bulletin board, and where such a wide variety of web bulletin boards are written (accessed) by a wide variety of writing users. From the viewpoint of who is writing more on the web bulletin board accepted by the layer, the importance of the web bulletin board and the writing user is calculated and the information is provided to the user 1. With this information, for example, the user 1 can specify a writing user who is the center of a writing user group in marketing, and can perform marketing activities such as sales promotion activities and advertising activities centering on the writing user. As a result, the user 1 can perform marketing activities more efficiently, for example, by increasing the amount and frequency of writing on the important WEB bulletin board by the important writing user.

  Of course, the content and the user may be tangible or intangible other than those described above. For example, the content may be an article, the user may be a purchaser of the article, the content may be an online site content, the user may be a download user of the content, the content may be a WEB bulletin board, a web log, a homepage, etc. Articles or comments (that is, information) posted (written) on the site may be used, and the user may be the user who wrote the information. Further, the content and the user may be physically the same type (for example, a human) such as a woman and a man and an adult and a child. Furthermore, content and users may be information, such as company sales and stock prices, temperature and humidity, and the like.

  The analysis support apparatus 10 includes an information processing unit 11, a user interface unit 12, a database interface unit 13, and a database 14.

  When the information processing unit 11 acquires the request of the user 1 via the user interface 12, the information processing unit 11 analyzes the importance of the user according to the request, generates information using the user importance, 12 to the user 1. That is, the information processing unit 11 can identify an important user that affects more users by analyzing the user importance of each user. At this time, the information processing unit 11 analyzes the mediation centrality of each content, that is, the degree of influence on others, and analyzes the user importance of each user using the analysis result, thereby more accurately. It is possible to identify an important user (calculate user importance). The user 1 can perform marketing activities more efficiently, for example, based on the analysis information thus obtained (information utilizing the user importance).

  The information processing unit 11 includes an importance level analysis unit 21 and a user importance level utilization information generation unit 22. The importance level analysis unit 21 performs processing related to analysis of the importance level of content, users, and the like. That is, the importance level analysis unit 21 analyzes the content to be targeted for marketing and the relationship structure of users, and generates information such as user importance level that can be used as a material for determining a marketing policy.

  The importance analysis unit 21 generates a content network 43 indicating the relationship between the contents in the content group, a network generation unit 31 that generates a user network 44 indicating the relationship between the users in the user group, and the user importance based on the networks. A user importance degree calculation unit 32 for calculating the degree is provided.

  The user importance level calculation unit 32 uses the content network 43 and the user network 44 generated by the network generation unit 31 in calculating the user importance level, which means that the information on the content is easily distributed via the user. It is possible to easily analyze the importance of the content and the user from the viewpoint. That is, the user importance level calculation unit 32 can easily calculate the user importance level of each user regarding the magnitude of influence on other users. Thereby, the user 1 can perform a marketing activity more efficiently, for example.

  The user importance utilization information generation unit 22 generates the user importance utilization information utilizing the user importance calculated by the importance analysis unit 21, and the user 1 handles the analysis result of the importance analysis unit 21, for example. The information is converted into easy-to-use information and supplied to the user 1. Thus, the user importance utilization information generation unit 22 can present the user importance to the user 1 as arbitrary information.

  The user interface unit 12 is a processing unit that performs interface processing with respect to the user 1. For example, the user interface unit 12 is generated by an input unit such as a keyboard or a mouse for the user 1 to input a request, or the user importance utilization information generation unit 22. An output unit such as a monitor or a speaker that presents the user importance utilization information to the user 1 as image information or audio information is provided.

  The database interface unit 13 performs interface processing of the database 14 with respect to the information processing unit 11. For example, the database interface unit 13 acquires various data stored in the database 14 based on an instruction from the information processing unit 11 (importance analysis unit 21), and supplies the data to the request source. Various data supplied from the unit 11 (importance analysis unit 21) is supplied to the database 14 and stored therein.

  The database 14 is a storage unit configured by a recording medium such as a hard disk or a flash memory, for example, a user log 41 which is log information on trends and states of user contents, content information 42 which is information on each content, and between contents Various information such as a content network 43 indicating the relationship between the user and a user network 44 indicating the relationship between the users is stored. Thereby, the information processing unit 11 can use various information such as the user log 41, the content information 42, the content network 43, and the user network 44 at an arbitrary timing.

  The user log 41 is information on each user, and also shows the relationship with each content. The content information 42 is information regarding each content. That is, a network having both the content and the user as nodes is constructed from the user log 41 and the content information 42. Based on this network, the network generation unit 31 generates a content network 43 having only contents as nodes and a user network 44 having only users as nodes.

  That is, the content network 43 and the user network 44 of the database 14 are information generated by the network generation unit 31 and may not be stored in the database 14 in the initial state.

  FIG. 2 is a block diagram illustrating a detailed configuration example of the network generation unit 31 in FIG.

  In FIG. 2, the network generation unit 31 includes a log analysis unit 51 that analyzes a user log 41, a content network generation unit 52 that generates a content network 43 based on log analysis results and content information 42, and the like. A user network generation unit 53 that generates the network 44 is included.

  The log analysis unit 51 acquires the user log 41 and the content information 42 stored in the database 14 via the database interface unit 13, analyzes the relationship between each user and each content from the user log 41, and the content information 42 By analyzing each content, a mixed network 45 is generated in which both the content and the user are nodes as shown in FIG. 3, for example, and the relationship between the content and the user is shown as a link connecting the nodes.

  In the mixed network 45 shown in FIG. 3, the contents C1 to C5 and the users u1 to u4 indicate nine nodes, and each line (link) connecting the contents and the user is represented by each content and each user. It shows the relationship between users. For example, if the contents C1 to C5 indicate products, and the users u1 to u4 are purchasers of the products, the link indicates the relationship between the product and the purchaser who purchased the product. That is, in that case, the mixed network 45 in FIG. 3 has the user u1 purchase the content C1 and the content C5, the user u2 purchases the content C1, the content C2, and the content C3, and the user u3 receives the content C3, Content C4 and content C5 are purchased, and user u4 indicates that content C4 and content C5 have been purchased. In other words, in the mixed network 45 of FIG. 3, the content C1 is purchased by the user u1 and the user u2, the content C2 is purchased by the user u2, the content C3 is purchased by the user u2 and the user u3, and the content C4 is purchased by the user u3. It is purchased by the user u4 and the content C5 is purchased by the user u1, the user u3, and the user u4.

  The log analysis unit 51 supplies the generated mixed network 45 to the content network generation unit 52 and the user network generation unit 53.

  When the content network generation unit 52 acquires such a mixed network 45, the content network generation unit 52 extracts only the relationship between the contents from the mixed network 45, and generates the content network 43 having only the content as a node as shown in FIG. To do.

  In the content network 43 shown in FIG. 4, content C1 to content C5 indicate five nodes extracted from the mixed network 45 in FIG. 3, and each line (link) connecting these nodes is the mixed node in FIG. The relationship between nodes in the network 45 is shown. That is, the content network 43 is a network that focuses only on the relationship between the contents in the mixed network 45 of FIG. 3 without being aware of the user's node.

  For example, in the mixed network 45 of FIG. 3, the content C1 is related to the content C2 and the content C3 via the user 2, and is related to the content C5 via the user u1. Therefore, in the content network 43 of FIG. 4, the content C1 is connected to the content C2, the content C3, and the content C5 by links. Similarly, since the content C2 is related to the content C1 and the content C3 via the user u2 in the mixed network 45 of FIG. 3, the content network 43 of FIG. 4 is linked to the content C1 and the content C3, respectively. Tied. Similarly, the content C3 is related to the content C1 and the content C2 through the user u2 and the content C4 and the content C5 through the user u3 in the mixed network 45 of FIG. In the content network 43, the content C1, the content C2, the content C4, and the content C5 are connected by links. Furthermore, since the content C4 is also related to the content C3 and the content C5 through the user u3 and the content C5 through the user u4 in the mixed network 45 of FIG. 3, the content network 43 of FIG. Are linked to the contents C3 and C5 by links. Similarly, the content C5 is related to the content C1 through the user u1, the content C3 and the content C4 through the user u3, and the content C4 through the user u4 in the mixed network 45 of FIG. Therefore, in the content network 43 of FIG. 4, the content C1, the content C3, and the content C4 are respectively connected by links.

  That is, for example, in the mixed network 45 of FIG. 3, if the contents C1 to C5 indicate products, and the users u1 to u4 are purchasers of the products, the links of the content network 43 are two products at both ends. Indicates that it was purchased by the same buyer.

  The content network generation unit 52 supplies the generated content network 43 to the database 14 via the database interface unit 13 and stores it.

  When the user network generation unit 53 obtains the mixed network 45 from the log analysis unit 51, the user network generation unit 53 extracts only the relationship between the users from the mixed network 45, and as shown in FIG. Is generated.

  In the user network 44 shown in FIG. 5, the users u1 to u4 indicate four nodes extracted from the mixed network 45 in FIG. 3, and each line (link) connecting these nodes is the mixed node in FIG. The relationship between nodes in the network 45 is shown. That is, the user network 44 is a network that focuses only on the relationship between users without being aware of the content node in the mixed network 45 of FIG.

  For example, in the mixed network 45 of FIG. 3, the user u1 is related to the user u2 via the content C1, and is related to the user u3 and the user u4 via the content C5. Therefore, in the user network 44 of FIG. 5, the user u1 is linked with the user u2, the user u3, and the user u4. Similarly, since the user u2 is related to the user u1 via the content C1 and the user u3 via the content C3 in the mixed network 45 shown in FIG. 3, the user u2 is related to the user u3 via the content C3. Each of u1 and user u3 is linked by a link. Similarly, the user u3 is related to the user u1 and the user u4 via the content C5, the user u2 via the content C3, and the user u4 via the content C4 in the mixed network 45 of FIG. Therefore, in the content network 43 of FIG. 4, the user u1, the user u2, and the user u4 are connected by links. Furthermore, since the user u4 is also related to the user u3 via the content C4 and the user u1 and the user u3 via the content C5 in the mixed network 45 shown in FIG. 3, the user network 44 shown in FIG. In FIG. 5, the user u1 and the user u3 are connected by links.

  That is, for example, in the mixed network 45 of FIG. 3, if the contents C1 to C5 indicate products, and the users u1 to u4 are purchasers of the products, the link of the user network 44 has two purchases at both ends. Indicates that the person has purchased the same product.

  The user network generation unit 53 supplies the generated user network 44 to the database 14 via the database interface unit 13 and stores it.

  Thus, since the network generation unit 31 generates the content network 43 and the user network 44 via the mixed network 45, the content network 43 and the user network 44 are easily generated from the user log 41 and the content information 42. be able to.

  FIG. 6 is a block diagram illustrating a detailed configuration example of the user importance degree calculation unit 32 of FIG.

  In FIG. 6, the user importance calculation unit 32 includes a content mediation centrality calculation unit 61, a user mediation centrality calculation unit 62, and a calculation unit 63.

  The content mediation centrality calculation unit 61 calculates the content mediation centrality of each content based on the content network 43. The content mediation centrality is a parameter indicating the magnitude of the ratio involved in the relationship between arbitrary contents in the content network 43.

  More specifically, in the content network 43, the content mediation centrality of the content A is that when the other two contents of the content A are connected by the shortest path along the link, the shortest path is the content A This is a parameter indicating the size of the rate of passing through. That is, the content mediation centrality is a parameter indicating the degree of influence on the connection between other contents.

  More specifically, for example, the content is a product and the user is a purchaser of the product, and the link between the content and the user in the mixed network 45 is a purchase relationship, that is, the purchaser and the purchaser purchases it. Assume that the product is connected to a product. In the content network 43 created based on such a mixed network 45, a strong connection between two products (a large number of paths, a short path length, etc.) means that, for example, purchase of purchasing the product This shows that the demographics are similar (close) to each other. Therefore, in this case, when marketing activities are performed on a purchaser of one product, it is highly possible that the marketing activity is highly influenced by the purchaser of the other product.

  On the other hand, when the connection between two products is weak (the number of passes is small, the pass length is long, etc.), for example, the buyers of each product are not similar (distant). Show. Therefore, in this case, even if the marketing activity is performed on the purchaser of one product, it is considered that the marketing activity does not easily affect the purchaser of the other product.

  The content mediation centrality is a parameter indicating the possibility that the presence or absence of the content causes a change in the strength of the connection with respect to the surrounding content. That is, as the content mediation centrality value is larger, the degree of influence is larger, and if the content is deleted from the content network 43, the strength of the connection between the surrounding content is more likely to change. Become.

  The change in the strength of the connection between contents may affect the marketing policy determined by the user 1. In other words, the content mediation centrality can also be said to be a parameter indicating the degree to which the presence or absence of the content affects the user 1. In other words, it can be said that the content with the larger content mediation centrality is more important content for the user 1.

The content mediation centrality value will be described more specifically. The content-mediated centrality Ce (n _i ) of a predetermined node n _i is calculated by the following equation (1).

In Expression (1), n indicates the number of combinations of two nodes other than the target node n _i , and n _st is a node s and a node t that are two nodes other than the node n _i. The number of paths with the shortest path length (shortest path) is shown, and g _i ^(st) indicates the number of paths that pass through the node n _i among n _st . However, the value s is smaller than the value t.

In other words, the node n _i content mediated centrality of Ce (n _i), of the shortest path between the nodes s and t, the ratio of the path passing through the node n _i, the combination of all the nodes s and t Average values are shown.

  The content mediation centrality calculation unit 61 calculates the content mediation centrality for all nodes (contents) using the above-described formula (1) based on the relationship of the content network 43, and calculates these values. 63.

  The user-mediated centrality calculation unit 62 performs the same processing as the content-mediated centrality calculation unit 61 on the user network 44 to calculate the user-mediated centrality of each user. The user-mediated centrality is a parameter indicating the magnitude of the ratio involved in the relationship between arbitrary users in the user network 44.

  Specifically, in the user network 44, the user-mediated centrality of the user B is that when the two other arbitrary users of the user B are connected by the shortest path along the link, the shortest path is the user B. This is a parameter indicating the size of the rate of passing through. That is, the user-mediated centrality is a parameter indicating the degree of influence on the connection between other users.

  More specifically, for example, the content is a product and the user is a purchaser of the product, and the link between the content and the user in the mixed network 45 is the purchase relationship, that is, the purchaser and the product purchased by the purchaser. Suppose that In the user network 44 created based on such a mixed network 45, a strong connection between two users (a large number of paths, a short path length, etc.) means that, for example, a product purchased by the user (Purchase pattern) are similar (close) to each other. Therefore, in this case, when marketing activities are performed using one purchaser as a model, it is likely that the other purchasers are likely to be greatly affected by the marketing activities. For example, if a discount service is performed in which the purchase pattern of one purchaser is the center of the discount target, the purchase pattern of the other purchaser who is strongly linked is likely to fall under the discount target.

  On the other hand, that the connection between two products is weak (the number of passes is small, the pass length is long, etc.) indicates that, for example, the purchase patterns of each other are not similar (distant). . Therefore, in this case, even if the marketing activity is performed with one purchaser as a model, it is considered that the marketing activity does not easily affect the other purchaser.

  The user-mediated centrality is a parameter indicating that the presence or absence of the user may cause a change in the strength of the connection with respect to surrounding users. In other words, the greater the user-mediated centrality value, the greater the degree of influence. If the user is deleted from the user network 44, there is a high possibility that the strength of the connection between the surrounding users will change more greatly. Become.

  The change in the strength of the connection between users may affect the marketing policy determined by the user 1. That is, the user-mediated centrality can also be said to be a parameter indicating the degree to which the presence or absence of the user affects the user 1. In other words, it can be said that a user having a larger user-mediated centrality value is an important user for the user 1.

The user-mediated centrality value will be described more specifically. The user-mediated centrality UC (n _i ) of the predetermined node n _i is calculated by the same formula as the formula (1) described above.

In other words, the node n user mediated centrality UC (n _i) also of _i, of the shortest path between the nodes s and t, the ratio of the path passing through the node n _i, the combination of all the nodes s and t Average values are shown.

  The user-mediated centrality calculation unit 62 calculates the user-mediated centrality for all nodes (users) using the above-described equation (1) based on the relationship of the user network 44, and calculates those values. 63.

  The calculation unit 63 performs a predetermined calculation based on the content-mediated centrality value supplied from the content-mediated centrality calculation unit 61 and the user-mediated centrality value supplied from the user-mediated centrality calculation unit 62. The user importance level indicating the importance level of each user based on the strength of influence on others is calculated.

The predetermined calculation will be described more specifically. When the content mediation centrality value is supplied from the content mediation centrality calculation unit 61, the computation unit 63 computes the following expression (2) and links each user in the mixed network 45 with a link to the user. The total CB (u _j ) of the content mediation centrality of all content to be calculated is calculated.

CB (u _j ) = ΣCe (n _ij ) (2)

In Expression (2), u _j represents the j-th user, and node n _ij represents a content node linked to the user u _j by a link.

Further, as shown in the following formula (3), the calculation unit 63 further calculates the value CB (u _j ) obtained by the formula (2) and the user-mediated centrality supplied from the user-mediated centrality calculation unit 62. Value UC (u _i ) is added to calculate the user importance IC (u _i ) for each user u _i .

IC (u _i ) = CB (u _i ) + UC (u _i ) (3)

  A specific example of calculating the user importance will be described below.

  For example, in the case of the content network 43 in FIG. 4, since the number of nodes is 5, n = 4. Considering the content C1, there are six combinations of nodes other than the content C1, C2-C3, C2-C4, C2-C5, C3-C4, C3-C5, and C4-C5.

Among these, the path (C2-C3, C3-C4, C3-C5, C4-C5) of a pair of adjacent nodes does not pass through the content C1, and therefore g _i ^(st) = 0 for these. . From FIG. 4, the shortest path of C2-C4 is one of content C2.fwdarw.content C3.fwdarw.content C4, and does not pass through the content C1, so that g _i ^(st) = 0. From FIG. 4, there are two shortest paths of C2-C5: content C2.fwdarw.content C1.fwdarw.content C5, content C2.fwdarw.content C3.fwdarw.content C5, and one of them passes through the content C1. _st = 2, g _i ^(st) = 1.

  Therefore, from the above equation (1), the content mediation centrality Ce (C1) of the content C1 is expressed by the following equation (4).

  Ce (C1) = (1/2) /6=0.083 (4)

  Next, considering the content C2, there are six combinations of nodes other than the content C2, C1-C3, C1-C4, C1-C5, C3-C4, C3-C5, and C4-C5.

Among these, the path (C1-C3, C1-C5, C3-C4, C3-C5, C4-C5) of the pair of adjacent nodes does not pass through the content C2, so g _i ^(st) = 0. Also, from FIG. 4, there are two shortest paths of C1 to C4: content C1 → content C3 → content C4 and content C1 → content C5 → content C4, but there is nothing that passes through the content C2. _i ^(st) = 0.

  Therefore, from the above-described equation (1), the content-mediated centrality Ce (C2) of the content C2 is expressed by the following equation (5).

  Ce (C2) = 0/6 = 0 (5)

  Next, considering the content C3, there are six combinations of nodes other than the content C3: C1-C2, C1-C4, C1-C5, C2-C4, C2-C5, and C4-C5.

Among these, the path (C1-C2, C1-C5, C4-C5) of the pair of adjacent nodes does not pass through the content C3, and therefore, g _i ^(st) = 0 for these. Further, from FIG. 4, there are two shortest paths of C1 to C4: content C1 → content C3 → content C4 and content C1 → content C5 → content C4, and one of them passes through the content C3. , N _st = 2 and g _i ^(st) = 1. Similarly, from FIG. 4, the shortest path of C2-C4 is one content C2-> content C3-> content C4, and one of the shortest paths passes through the content C3, so that n _st = 1, g _i ^(st) = 1.

  Therefore, from the above equation (1), the content mediation centrality Ce (C3) of the content C3 is expressed by the following equation (6).

  Ce (C3) = (0.5 + 1 + 0.5) /6=0.333 (6)

  Next, considering the content C4, there are six combinations of nodes other than the content C4: C1-C2, C1-C3, C1-C5, C2-C3, C2-C5, and C3-C5.

Of these, the path of the adjacent node pair (C1-C2, C1-C3 , C1-C5, C2-C3, C3-C5) , since no via content C4, for these, g _i ^(st) = 0. Also, from FIG. 4, there are two shortest paths of C2 to C5: content C2 → content C1 → content C5 and content C2 → content C3 → content C5, but there is nothing that passes through the content C4. _i ^(st) = 0.

  Therefore, from the above equation (1), the content mediation centrality Ce (C4) of the content C4 is expressed by the following equation (7).

  Ce (C4) = 0/6 = 0 (7)

  Next, considering the content C5, there are six combinations of nodes other than the content C5: C1-C2, C1-C3, C1-C4, C2-C3, C2-C4, and C3-C4.

Among these, the path (C1-C2, C1-C3, C2-C3, C3-C4) of the pair of adjacent nodes does not pass through the content C5, so that g _i ^(st) = 0 for these. . From FIG. 4, the shortest path of C2-C4 is one of content C2-> content C3-> content C4 and does not pass through the content C5, so g _i ^(st) = 0. From FIG. 4, there are two shortest paths of C1-C4: content C1 → content C3 → content C4 and content C1 → content C5 → content C4, and one of them passes through the content C5. n _st = 2 and g _i ^(st) = 1.

  Therefore, from the above equation (1), the content mediation centrality Ce (C5) of the content C5 is expressed by the following equation (8).

  Ce (C5) = (1/2) /6=0.083 (8)

From Equation (2) and Equations (4) to (8), the total value CB (u _j ) of content mediation centrality for each user is as shown in Equations (9) to (12) below.

CB (u1) = Ce (C1) + Ce (C5) = 0.166 (9)
CB (u2) = Ce (C1) + Ce (C2) + Ce (C3) = 0.416 (10)
CB (u3) = Ce (C3) + Ce (C4) + Ce (C5) = 0.416 (11)
CB (u4) = Ce (C4) + Ce (C5) = 0.083 (12)

  Considering the user u1, there are three combinations of nodes other than the user u1, u2-u3, u2-u4, u3-u4.

Among these, the path (u2-u3, u3-u4) of the pair of adjacent nodes does not pass through the user u1, and therefore, g _i ^(st) = 0. Further, from FIG. 5, there are two shortest paths of u2-u4: user u2 → user u1 → user u4 and user u2 → user u3 → user u4, and one of them passes through user u1. , N _st = 2 and g _i ^(st) = 1.

  Therefore, from the above equation (1), the user-mediated centrality UC (u1) of the user u1 is represented by the following equation (13).

  UC (u1) = (1/2) /3=0.167 (13)

  Next, considering the user u2, there are three combinations of nodes other than the user u2: u1-u3, u1-u4, u3-u4.

Since these nodes are all adjacent to each other and do not pass through the user u2, g _i ^(st) = 0.

  Therefore, from the above equation (1), the user-mediated centrality UC (u2) of the user u2 is represented by the following equation (14).

  UC (u2) = 0/3 = 0 (14)

  Next, regarding the user u3, there are three combinations of nodes other than the user u3: u1-u2, u1-u4, u2-u4.

Among these, since the path (u1-u2, u1-u4) of the pair of adjacent nodes does not pass through the user u3, g _i ^(st) = 0 for these. Further, there are two shortest paths of u2-u4 from user u2 → user u1 → user u4 and user u2 → user u3 → user u4, and one of them passes through user u3. , N _st = 2 and g _i ^(st) = 1.

  Therefore, from the above equation (1), the user-mediated centrality UC (u3) of the user u3 is represented by the following equation (15).

  UC (u3) = (1/2) /3=0.167 (15)

  Next, regarding the user u4, there are three combinations of nodes other than the user u4: u1-u2, u1-u3, u2-u3.

Since these nodes are all adjacent to each other and do not pass through the user u4, g _i ^(st) = 0.

  Therefore, from the above equation (1), the user-mediated centrality UC (u4) of the user u4 is represented by the following equation (16).

  UC (u4) = 0/3 = 0 (16)

Based on Expression (3), Expression (13) to Expression (16), the user importance IC (u _i ) of each user is represented by the following Expression (17) to Expression (20).

IC (u1) = CB (u1) + UC (u1) = 0.333 (17)
IC (u2) = CB (u2) + UC (u2) = 0.416 (18)
IC (u3) = CB (u3) + UC (u3) = 0.583 (19)
IC (u4) = CB (u4) + UC (u4) = 0.083 (20)

  That is, in this case, the user importance of the user u3 is the highest. The user importance calculation unit 32 in FIG. 1 calculates the user importance as described above for each user, and supplies it to the user importance utilization information generation unit 22.

  Next, an example of a specific flow of the above process will be described.

  First, an example of the flow of user importance utilization processing executed by the information processing unit 11 of FIG. 1 will be described with reference to the flowchart of FIG.

  When the user importance utilization process is started, the importance analysis unit 21 of the information processing unit 11 performs a process related to the importance analysis in step S1. Details of the importance analysis processing will be described later.

  In step S2, the user importance utilization information generation unit 22 generates user importance utilization information utilizing the user importance based on the user importance obtained by the process of step S1. The generated user importance use information is presented to the user 1 via the user interface unit 12.

  When the process of step S2 ends, the information processing unit 11 ends the user importance utilization process.

  Next, an example of the flow of importance analysis processing executed in step S1 of FIG. 7 will be described with reference to the flowchart of FIG.

  In step S21, the network generation unit 31 of the importance analysis unit 21 generates a network (for example, a mixed network 45, a content network 43, a user network 44, etc.) necessary for distribution analysis. A specific example of the network generation process will be described later.

  When the process of step S21 ends, the user importance level calculation unit 32 calculates the user importance level in step S22. A specific example of the user importance calculation process will be described later.

  When the calculation of the user importance level is completed, the importance level analysis unit 21 ends the importance level analysis process, returns the process to step S1 in FIG. 7, and executes the subsequent processes.

  Next, an example of the flow of network generation processing executed in step S21 of FIG. 8 will be described with reference to the flowchart of FIG.

  When the network generation process is started, in step S41, the log analysis unit 51 of the network generation unit 31 analyzes the user log 41 using the content information 42 and generates, for example, a mixed network 45.

  In step S42, the content network generation unit 52 generates the content network 43 based on the log analysis result (for example, the mixed network 45). In step S43, the user network generation unit 53 generates the user network 44 based on the log analysis result.

  When the process of step S43 ends, the network generation unit 31 ends the network generation process, returns the process to step S21 of FIG. 8, and executes the subsequent processes.

  Next, an example of the flow of user importance calculation processing executed in step S22 of FIG. 8 will be described with reference to the flowchart of FIG.

  When the user importance calculation process is started, the content mediation centrality calculation unit 61 calculates content mediation centrality based on the content network 43 in step S61.

  In step S62, the calculation unit 63 calculates the total content mediation centrality of the content held by each user.

  In step S63, the user-mediated centrality calculation unit 62 calculates the user-mediated centrality.

  In step S64, the calculation unit 63 calculates the user importance of each user based on the total content mediation centrality for each user calculated in step S62 and the user mediation centrality calculated in step S63. .

  When the process of step S64 ends, the user importance level calculation unit 32 ends the user importance level calculation process, returns the process to step S22 of FIG. 8, and ends the importance level analysis process.

  As described above, when each process is executed and the user importance level is calculated, the user importance level utilization information generation unit 22 performs, for example, a user list sorted in order of the user importance level based on the supplied user importance level. Then, information utilizing the user importance, such as a user distribution map by user importance, marketing method recommendation information based on the user importance, and the like are created and presented to the user 1 via the user interface unit 12.

  Thus, the analysis support apparatus 1 can analyze information. In particular, the magnitude of the influence of each node on other nodes in the network can be analyzed more accurately.

  The user 1 can analyze the magnitude of the influence of each node on other nodes in the network more accurately by referring to the presented user importance utilization information. Based on this, it is possible to take more effective measures such as marketing activities.

  In the above description, the analysis support apparatus 1 has been described so as to generate and provide the user importance utilization information. However, the analysis support apparatus 1 may calculate the importance of the user and the content as described above, For example, information utilizing the content importance may be provided together with the user importance utilization information, or only information utilizing the content importance may be provided, or the user importance and content Information using both importance levels may be provided.

  FIG. 11 is a diagram illustrating another example of the information processing unit in FIG.

  An information processing unit 111 illustrated in FIG. 11A further includes a content importance utilization information generation unit 122 in addition to the configuration of the information processing unit 11 illustrated in FIG. The content importance utilization information generation unit 122 uses the content mediation centrality calculated by the user importance calculation unit 32 as the content importance of each content, generates content importance utilization information that utilizes the content importance, It is presented to the user 1 via the user interface unit 12. That is, the information processing unit 111 generates content importance utilization information together with user importance utilization information and presents it to the user 1.

  By doing in this way, the analysis support apparatus 1 can provide the user 1 with more diverse information. In addition, since the analysis support apparatus 1 provides a plurality of types of information, the user 1 can select each information as necessary to perform more appropriate analysis.

  An information processing unit 211 illustrated in FIG. 11B includes an importance level analysis unit 221 and a content importance level utilization information generation unit 122 that obtain only the importance level of content, unlike the information processing unit 11 illustrated in FIG.

  The importance level analysis unit 221 includes the content network generation unit 52 in FIG. 2 and the content mediation centrality calculation unit 61 in FIG. That is, the importance analysis unit 221 generates only the content network 43 in the content network generation unit 52, and calculates only the mediation centrality of each content in the content mediation centrality calculation unit 61.

  The content importance utilization information 122 uses the content mediation centrality calculated in this way as the content importance of each content, generates content importance utilization information utilizing the content importance, and uses the content importance utilization information as a user interface unit. 12 to the user 1. That is, the information processing unit 211 generates only content importance utilization information and presents it to the user 1.

  By doing in this way, the analysis support apparatus 1 can generate | occur | produce content importance utilization information more easily, and can provide it to the user 1. FIG. That is, since the information processing unit 211 omits the generation of the user importance utilization information, the information importance utilization information is generated at a lower load and faster than the information processing unit 111 in FIG. 1 can be provided. Accordingly, the user 1 who does not need the user importance utilization information can acquire the content importance utilization information at a higher speed, and thereby can perform an appropriate analysis.

  The information processing unit 311 illustrated in FIG. 11C includes an importance level analysis unit 21 and a user importance level / content importance level utilization information generation unit 322. The user importance / content importance utilization information generation unit 322 utilizes both the user importance and the content importance (content mediation centrality) calculated by the user importance calculation unit 32 of the importance analysis unit 21. The degree / content importance utilization information is generated and presented to the user 1 via the user interface unit 12. That is, the information processing unit 311 generates user importance / content importance utilization information and presents it to the user 1.

  By doing in this way, the analysis support apparatus 1 can provide the user 1 with more diverse information. In this case, the analysis support apparatus 1 presents the user importance level / content importance level utilization information using both the user importance level and the content importance level (content mediation centrality) to the user 1, which is more sophisticated and complicated. A simple analysis result can be presented to the user 1. Thereby, the user 1 may be able to perform advanced analysis more easily than the information processing unit 111 in FIG. 11A.

  Regarding these examples, a specific processing flow will be described.

  With reference to the flowchart of FIG. 12, an example of the flow of the user importance / content importance utilization processing executed by the information processing unit 111 of FIG. 11A will be described.

  When the user importance level / content importance level utilization process is started, in step S101, the importance level analysis unit 21 executes the importance level analysis process shown in FIG. 8 as in the case of step S1 in FIG. Analyzes user importance and content importance (content mediation centrality). In step S102, the user importance utilization information generation unit 22 generates user importance utilization information that utilizes the user importance calculated in the process of step S101, as in step S2 of FIG. In step S103, the content importance utilization information generation unit 122 sets the content mediation centrality calculated in the process of step S101 as the content importance, and generates content importance utilization information that uses the content importance. When the process of step S103 is completed, the information processing unit 111 ends the user importance / content importance utilization process.

  Next, an example of the flow of content importance utilization processing executed by the information processing unit 211 in FIG. 11B will be described with reference to the flowchart in FIG.

  When the content importance use process is started, in step S121, the content network generation unit 52 of the importance analysis unit 221 generates the content network 43 as in the case of step S42 in FIG. In S122, the content mediation centrality calculation unit 61 calculates content mediation centrality as in the case of step S61 in FIG. In step S123, the content importance utilization information generation unit 122 sets the content mediation centrality calculated in the process of step S122 as the content importance as in the case of step S103 in FIG. 12, and uses the content importance. Generate usage information. When the process of step S123 ends, the information processing unit 211 ends the content importance utilization process.

  Next, another example of the user importance level / content importance level utilization process executed by the information processing unit 311 in FIG. 11C will be described with reference to the flowchart in FIG. 14.

  In step S141, as in the case of step S1 in FIG. 7, the importance analysis unit 21 executes the importance analysis processing shown in FIG. 8, and analyzes the user importance and content importance (content mediated centrality). To do. In S142, the user importance level / content importance level utilization information generation unit 322 uses both the user importance level and the content importance level (content mediation centrality) calculated in the process of step S141. Generate usage information. When the process of step S142 is completed, the information processing unit 311 ends the user importance / content importance utilization process.

  In the user importance calculation process, the content mediation centrality calculation unit 61 may calculate the content mediation centrality in consideration of a weighting factor w based on an arbitrary index of each content. Any index may be used as an index of the weighting coefficient, for example, the selling price or popularity (reciprocal of popularity ranking) of content.

For example, when the weighting factor of the node n _s is set to w _s and the weighting factor of the node n _t is set to w _t , the content mediation centrality calculation unit 61 uses the following formula (21) instead of the formula (1). To calculate the content mediation centrality.

However, in Expression (21), as in Expression (1), n indicates the number of combinations of two nodes other than the target node n _i , and n _st is other than the node n _i . The number of paths with the shortest path length (shortest path) between the two nodes s and t is shown, and g _i ^(st) is a path that passes through the node n _i out of n _st The number of The value s is smaller than the value t. Furthermore, w _st indicates the sum of w _s and w _t .

As shown in the equation (21), when determining the content-mediated centrality Ce (n _i ), the content-mediated centrality calculating unit 61 uses g _i ^(st) / n _st as a parameter based on the content weighting factor. (Weight coefficient) w _st may be multiplied.

  An example of the flow of the user importance calculation process in this case will be described with reference to the flowchart of FIG.

  In step S <b> 161, the content mediation centrality calculation unit 61 calculates content mediation centrality in consideration of the weight coefficient of each content based on the content network 43. In step S162, the computing unit 63 calculates the total content mediation centrality of the content held by each user. In step S163, the user-mediated centrality calculation unit 62 calculates the user-mediated centrality. In step S164, the calculation unit 63 calculates the user importance level. When the process of step S164 ends, the user importance level calculation unit 32 ends the user importance level calculation process, returns the process to step S22 of FIG. 8, and ends the importance level analysis process.

  By doing in this way, the content mediation centrality calculation unit 61 can reflect the criteria other than the above-described content circulation (influence on other content) in the content mediation centrality. That is, the content mediation centrality calculation unit 61 can generate parameters based on various values.

  It should be noted that, as in the case of calculating the content-mediated centrality, it is of course possible to consider each user's weighting factor when calculating the user-mediated centrality.

  In the above description, the information processing unit 11 that calculates the user importance is described as being configured as a part of the analysis support apparatus 1 that presents the user importance utilization information to the user 1. The information processing unit 11 may be applied to any device.

  FIG. 16 is a block diagram showing a configuration example of a network system to which the present invention is applied.

  In FIG. 16, the network system includes a content distribution server 411, a terminal device 412, and a terminal device 413 connected to the network 410.

  The content distribution server 411 is a server for distributing content, a content distribution service unit 421 that performs processing related to content distribution, a log creation unit 422 that generates log information of content distribution processing, and promotion of sales (distribution) of content. It has a sales promotion processing unit 423, an information processing unit 11, a database interface unit 13, and a database 14 that perform processing.

  That is, the content distribution server 411 includes a content distribution service unit 421, a log creation unit 422, and a sales promotion processing unit 423 instead of the user interface 12 of the analysis support apparatus 10.

  The content distribution service unit 421 is a processing unit that provides a content distribution service to the terminal device 412 and the terminal device 413 via the network 410 based on those requests. The content distribution service unit 421 supplies a request from the terminal device 412 or the terminal device 413 and information such as processing contents for the request to the log creating unit 422. Based on the information, the log creation unit 422 creates a user log 41 that is log information regarding the content distributed with the terminal device (or the user), and stores the user log 41 in the database 14. The content information 42 may be stored in the database 14 in advance, or may be registered in the database 14 by the log creation unit 422 together with the user log 41.

  For example, the sales promotion processing unit 423 creates recommendation information for the content to be distributed by the content distribution service unit 421, and accepts the content distribution request published to the network 410 by the content distribution service unit 421. A discount service and sample data are provided to the terminal device 412 and the terminal device 413 with high user importance. At that time, the sales promotion processing unit 423 performs those sales promotion activities based on the user importance utilization information generated by the information processing unit 11.

  By doing in this way, the sales promotion process part 423 can perform more effective sales promotion activity.

  In the above description, the types of nodes (number of parameters) in the network have been described as being composed of two types (two parameters) of content and user. However, there are three types of nodes (number of parameters). It may be the above. In this case, the information processing unit 11 may be able to calculate the mediation centrality for each type.

  In the above, expressions for calculating mediation centrality, user importance, and the like have been described. However, these are all examples, and the information processing unit uses contents other than those described above to determine content mediation centrality and user importance. Mediation centrality, user importance, content importance, and the like may be calculated.

  Further, in the above description, the user 1 has been described as determining the marketing policy based on the user importance utilization information. However, the use of the user importance utilization information is not limited to marketing. It may be used for such a purpose.

  The series of processes described above can be executed by hardware or can be executed by software. In this case, for example, it may be configured as a personal computer as shown in FIG.

  In FIG. 17, a CPU (Central Processing Unit) 501 of the personal computer 500 performs various processes according to a program stored in a ROM (Read Only Memory) 502 or a program loaded from a storage unit 513 to a RAM (Random Access Memory) 503. Execute the process. The RAM 503 also appropriately stores data necessary for the CPU 501 to execute various processes.

  The CPU 501, ROM 502, and RAM 503 are connected to each other via a bus 504. An input / output interface 510 is also connected to the bus 504.

  The input / output interface 510 includes an input unit 511 including a keyboard and a mouse, a display including a CRT (Cathode Ray Tube) and an LCD (Liquid Crystal Display), an output unit 512 including a speaker, and a hard disk. A communication unit 514 including a storage unit 513 and a modem is connected. The communication unit 514 performs communication processing via a network including the Internet.

  A drive 515 is connected to the input / output interface 510 as necessary, and a removable medium 521 such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory is appropriately mounted, and a computer program read from them is It is installed in the storage unit 513 as necessary.

  When the above-described series of processing is executed by software, a program constituting the software is installed from a network or a recording medium.

  For example, as shown in FIG. 17, the recording medium is distributed to distribute the program to the user separately from the apparatus main body, and includes a magnetic disk (including a flexible disk) on which the program is recorded, an optical disk ( Removable media 521 including a CD-ROM (compact disk-read only memory), a DVD (digital versatile disk), a magneto-optical disk (including MD (mini-disk) (registered trademark)), or a semiconductor memory In addition to being configured, it is configured by a ROM 502 on which a program is recorded and a hard disk included in the storage unit 513, which is distributed to the user in a state of being pre-installed in the apparatus body.

  In the present specification, the step of describing the program recorded on the recording medium is not limited to the processing performed in chronological order according to the described order, but is not necessarily performed in chronological order. It also includes processes that are executed individually.

  Further, in this specification, the system represents the entire apparatus composed of a plurality of devices (apparatuses).

  In the above description, the configuration described as one device may be divided and configured as a plurality of devices. Conversely, the configurations described above as a plurality of devices may be combined into a single device. Of course, configurations other than those described above may be added to the configuration of each device. Furthermore, if the configuration and operation of the entire system are substantially the same, a part of the configuration of a certain device may be included in the configuration of another device. That is, the embodiment of the present invention is not limited to the above-described embodiment, and various modifications can be made without departing from the gist of the present invention.

  The present invention can be applied to an information processing apparatus.

It is a block diagram which shows the structural example of the analysis assistance apparatus to which this invention is applied. It is a block diagram which shows the detailed structural example of the network production | generation part of FIG. It is a figure which shows the example of the mixed network which uses a content and a user as a node. It is a figure which shows the example of the content network which uses the content of FIG. 3 as a node. It is a figure which shows the example of the user network which makes the user of FIG. 3 a node. It is a block diagram which shows the detailed structural example of the user importance calculation part of FIG. It is a flowchart explaining the example of the flow of a user importance utilization process. It is a flowchart explaining the example of the flow of an importance analysis process. It is a flowchart explaining the example of the flow of a network generation process. It is a flowchart explaining the example of the flow of a user importance calculation process. It is a figure which shows the other example of the information processing part of FIG. It is a flowchart explaining the example of the flow of a user importance and content importance utilization process. It is a flowchart explaining the example of the flow of a content importance utilization process. It is a flowchart explaining the example of the flow of a user importance and content importance utilization process. It is a flowchart explaining the other example of the flow of a user importance calculation process. It is a block diagram which shows the structural example of the network system to which this invention is applied. It is a block diagram which shows the structural example of the personal computer to which this invention is applied.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 Analysis support apparatus, 11 Information processing part, 21 Importance analysis part, 22 User importance utilization information generation part, 31 Network generation part, 32 User importance calculation part, 51 Log analysis part, 52 Content network generation part, 53 User Network generation unit 61 content mediation centrality calculation unit 62 user mediation centrality calculation unit 63 calculation unit 111 information processing unit 122 content importance utilization information generation unit 211 information processing unit 221 importance analysis unit 231 Content Mediation Centrality Calculation Unit, 311 Information Processing Unit, 322 User Importance / Content Importance Utilization Information Generation Unit, 411 Content Distribution Server, 421 Content Distribution Service Unit, 422 Log Creation Unit, 423 Sales Promotion Processing Unit

Claims (10)

A first parameter having only the first parameter as a node from a mixed network in which both the first parameter and the second parameter are nodes and the relationship between the first parameter and the second parameter is shown as a link connecting the nodes. Network generating means for generating a network and a second parameter network each having only the second parameter as a node;
A second parameter importance level for calculating the importance level of the second parameter using the first parameter network and the second parameter network generated by the network generation unit as indicators of the strength of influence on others. An information processing apparatus comprising: calculation means. The second parameter importance calculating means includes:
First parameter mediated centrality calculating means for calculating a first parameter mediated centrality indicating the strength of influence on the other of the first parameter based on the first parameter network generated by the network generating means; ,
Second parameter mediated centrality calculating means for calculating a second parameter mediated centrality indicating the strength of influence of the second parameter on the other based on the second parameter network generated by the network generating means; ,
Based on the first parameter-mediated centrality calculated by the first parameter-mediated centrality calculating means and the second parameter-mediated centrality calculated by the second parameter-mediated centrality calculating means, a predetermined calculation is performed. The information processing apparatus according to claim 1, further comprising: calculating means for calculating importance of the second parameter. The first parameter mediated centrality calculating means obtains a ratio of paths passing through the target node in a shortest path between the other two nodes of the target node in the first parameter network, and calculates the ratio as the other Calculating the first parameter-mediated centrality by averaging over all combinations of the two nodes of
The second parameter mediated centrality calculating means obtains a ratio of paths passing through the target node in a shortest path between the other two nodes of the target node in the second parameter network, and calculates the ratio as the other Calculating the second parameter-mediated centrality by averaging over all combinations of the two nodes of
The calculation means calculates a sum of the first parameter mediated centrality calculated by the first parameter mediated centrality calculation means of the first parameter corresponding to the second parameter in the mixed network, and the total sum The information processing apparatus according to claim 2, wherein an operation for calculating the sum of the second parameter-mediated centralities calculated by the second parameter-mediated centrality calculating unit is performed for each second parameter. In the first parameter network, the first parameter mediated centrality calculating means calculates the ratio of the path passing through the target node among the shortest paths between the other two nodes of the target node. The information processing apparatus according to claim 3, wherein the first parameter is multiplied by a weighting factor of an index different from the first parameter-mediated centrality. First parameter related information related to the first parameter, second parameter related information related to the second parameter, information for generating the mixed network, and the first parameter network and the second parameter network. The information processing apparatus according to claim 1, further comprising storage means for storing. The information processing according to claim 1, further comprising: second parameter importance utilization information generation means for generating second parameter utilization information utilizing the importance of the second parameter calculated by the second parameter importance calculation means. apparatus. The second parameter importance calculating means further calculates the importance of the first parameter using the strength of influence on others as an index based on the first parameter network generated by the network generating means;
The information processing according to claim 1, further comprising first parameter importance utilization information generation means for generating first parameter utilization information utilizing the importance of the first parameter calculated by the second parameter importance calculation means. apparatus. A first parameter having only the first parameter as a node from a mixed network in which both the first parameter and the second parameter are nodes and the relationship between the first parameter and the second parameter is shown as a link connecting the nodes. A network and a second parameter network each having only the second parameter as a node;
An information processing method for executing the step of calculating the importance of the second parameter using the first parameter network and the second parameter network as indices of the strength of influence on others. A first parameter having only the first parameter as a node from a mixed network in which both the first parameter and the second parameter are nodes and the relationship between the first parameter and the second parameter is shown as a link connecting the nodes. A network and a second parameter network each having only the second parameter as a node;
A program for causing a computer to execute the step of calculating the importance of the second parameter using the first parameter network and the second parameter network as an index of the strength of influence on others.   A recording medium on which the program according to claim 9 is recorded.

Images