JP2010146415A - Server for providing online service - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To share the information of a social graph among respective web services. <P>SOLUTION: In a web service X, an irreversible operation is executed to the identification tag of each user, and the social graph in the web service X is converted to the anonymous social graph of the already operated identification tag and is transmitted to the server of a network service N. In the web service Y and the web service Z, the respective anonymous social graphs are also transmitted to the server of the network service N by similar processing. In the server of the network service N, the anonymous social graphs obtained from the respective web services are synthesized to generate a synthesis anonymous social graph. The server of the network service N transmits the already operated identification tag of an acquaintance or an indirect acquaintance in the other services to the respective web services from the synthesis anonymous social graph. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

The present invention relates to information processing in which a plurality of servers cooperate to provide an online service.

In recent years, a community-type Web service called a social networking site (SNS) that builds and promotes a connection between users on the Internet has been widely spread (for example, see Patent Document 1). For example, a website disclosed in Non-Patent Document 1 is known as an SNS.

Recently, the service is not limited to SNS, but is a service similar to SNS, and services on the Web having a function of managing the connection between users are rapidly increasing. The same as SNS for the purpose of building and promoting the connection between users in the provision of all services provided on the Internet for the purpose of searching, sharing photos and videos, selling, chatting and exchanging messages by voice. Function is provided. In the present specification, hereinafter, a service on the Web having a function of managing the connection between users regardless of the main purpose of the service is referred to as a “Web service”. Currently, most users use a plurality of Web services. Each Web service can have different acquaintances, communicate in different ways, and disclose different information. In this situation, the user is considered to have the following two requirements.

The first request is a request for a unified interface across a plurality of Web services. As a result of most users using a plurality of Web services, information management for each Web service is difficult. There is no way for a user to collectively manage who can communicate with which Web service and which information is disclosed to whom with which Web service. In such a situation, a service that can handle a plurality of Web services in a unified manner is desired.

On the other hand, a framework for providing a unified interface to users by interchanging information handled between Web services has been proposed (for example, see Non-Patent Document 2). This is because each Web service publishes a unified API (Application Program Interface) so that a list of acquaintances inside the Web service, acquaintances, information registered by each user can be obtained from an external server. Standards are used to make However, with such a method, there is a possibility that information previously held by each Web service provider may be leaked to the outside. This is not preferable for most Web service providers mainly from a business viewpoint. Accordingly, the use of information interchange through public APIs is limited to some SNSs, and is not widely used for Web services in general.

The second request is a request to become a new user and acquaintance more effectively across a plurality of Web services and to expand the human network. For example, people who post diaries or exchange messages on the SNS service may want to be acquainted with each other on a chat or voice call web service. Or, if you can get acquainted with an acquaintance of an acquaintance, that is, an indirect acquaintance, you may be able to expand your relationship.

In order for acquaintances to get to know each other through a Web service, a certain user usually inputs and registers various information about himself / herself. Another user searches for information inputted by the user, thereby discovering a candidate (acquaintance candidate) of a user who becomes an acquaintance on the Web service, and constructs a new human relationship. For example, each user registers the things he / she is impressed with, the place of birth, place of residence, school of origin, etc. in the Web service server, and each user searches for them to find an acquaintance candidate . Some Web services use real names or e-mail addresses as search targets. Such Web services have high real names. In addition, there is a Web service that searches for acquaintance candidates only with highly anonymous information such as nicknames and things that have an impression. Usually, such an acquaintance candidate discovery is performed by a so-called active search operation in which a user actively transmits a search term to a Web service.

On the other hand, it is conceivable to provide an acquaintance candidate recommendation service that does not involve an active search operation by the user. The following two methods are conceivable as conventional methods for selecting an acquaintance candidate in an acquaintance candidate recommendation service. The first method is a method of automatically selecting other users whose attribute information for each user input in advance is the same or similar as a candidate for acquaintance and recommending them. For example, when a user registers soccer as a matter of interest in a web service, another user who also registers soccer as a matter of interest in the web service is recommended as an acquaintance candidate (for example, non-patent) Reference 3).

As a second method of selecting an acquaintance candidate in an automatic acquaintance candidate recommendation service, a method of recommending a user who has a common acquaintance in the Web service as an acquaintance candidate is known (for example, (Refer nonpatent literature 4.). For example, in Web service X, user A and user B have an acquaintance relationship, and user B and user C have an acquaintance relationship, but user A and user C have no acquaintance relationship. At this time, user A and user C are more likely to become acquaintances than other users unrelated to users A, B, and C. Therefore, the Web service X recommends the user C to the user A and the user A to the user C as acquaintance candidates.
US Patent No. 7,069,308 http://www.myspace.com http://code.google.com/intl/en/apis/opensocial/ http://vkontakte.ru/ http://www.facebook.com/ http://www.foaf-project.org/

The problem to be solved by the present invention will be described with reference to FIG. FIG. 1 shows a user at a black dot node, an acquaintance relationship in the Web service X by a solid line edge, and an acquaintance relationship in the Web service Y by a dashed edge. For example, in FIG. 1, a user A 101 having an identification tag IDxa on the Web service X and a user B 102 having IDxb are in an acquaintance relationship 111 on the Web service X. Similarly, the user A having the identification tag IDya on the Web service Y is in an acquaintance relationship 112 with the user D having IDyd. In the present specification, an identification tag for uniquely identifying the user T in the Web service S is expressed as IDst. For example, IDxa shown in FIG. 1 is an identification tag for uniquely recognizing the user A by the Web service X. In FIG. 1, user A uses two services, Web service X and Web service Y. User B is an acquaintance on Web service X, and user D is an acquaintance on Web service Y. There are mainly two problems to be solved by the present invention.

The first problem is that it is difficult for a user to manage an acquaintance beyond the framework of the Web service while maintaining the independence of each Web service. In the example of FIG. 1, it becomes more convenient for user A to be able to handle user B and user D who are acquainted with each other on different Web services. For example, consider that user A browses information on user D on Web service X. For this purpose, it is necessary for the Web service X server to acquire information on the Web service Y. However, the acquaintance relationship in the Web service is regarded as a source of revenue for the operator of each Web service. For this reason, the operator of Web service Y tends to hate that it leaks to the server of Web service X. In the conventional method, due to the independence of each Web service, it has been difficult to realize flexible accommodation of user information beyond the Web service.

The second problem is that it is difficult to realize an acquaintance candidate recommendation service that exceeds the framework of the Web service while maintaining the independence of the Web services. As described above, user A has a demand to establish a new relationship with a new acquaintance beyond the framework of the Web service. At this time, a service for recommending an acquaintance of an acquaintance as a new acquaintance candidate has already been realized. For example, in FIG. 1, both user E and user C are indirect acquaintances of user A. Here, there is one common acquaintance between user A and user E, and two common acquaintances between user A and user C. Therefore, for user A, user C is considered to be a more important acquaintance candidate than user E. However, the fact that there are two common acquaintances between the user A and the user C cannot be detected by either the Web service X operator or the Web service Y operator. The fact that there are two common acquaintances between user A and user C is that the server of Web service X obtains information about the user of Web service Y, or vice versa, or a third party. A certain web service M needs to collect and analyze information about both users of the web services X and Y. However, the acquaintance relationship in the Web service is regarded as a source of revenue for the operator of each Web service. For this reason, Web service providers tend to hate that their acquaintance relationships are leaked to servers operated by other providers. In the conventional method, due to the independence of each Web service, accurate recommendation of acquaintance candidates beyond the Web service frame has been difficult or impossible.

In reality, most Internet users use a plurality of Web services and have different acquaintance relationships as illustrated in FIG. 1 for each Web service. However, in the above-described automatic acquaintance candidate recommendation service, the attribute information and acquaintance-related information used for candidate selection are limited to information stored in one Web service server. If attribute information and acquaintance relationships in a plurality of Web services can be used for candidate selection in an acquaintance candidate recommendation service, an acquaintance candidate who recommends more acquaintance candidates and has more common acquaintances in each Web service. Can be preferentially recommended. However, it has been impossible to provide such a Web service by the conventional method.

The present invention has been made in view of such a situation, and it is possible to integrate and handle attribute information and acquaintance relationships of a plurality of Web services without disclosing information held by each Web service to the outside. And a method for realizing a broader and more accurate acquaintance candidate recommendation service.

In one embodiment of the present invention, the first Web service is included in a first identification tag for uniquely identifying a user who uses the first Web service among the users of the first Web service. A first computed recognition card which has been computed by the first server, and a second for uniquely identifying the user using the second Web service among the users of the second Web service Receiving means for receiving, from the first server, a second computed recognition tag obtained by performing the computation on the first recognition server, and the second Web service including the second Web service. Receiving means for receiving, from the second server, a third computed recognition tag that has been subjected to the computation by the second server, and using the third server as a search key, the first server and Search for the computed identification tag received from the second server, Search means for detecting a match between the second computed recognition tag and the third computed recognition tag, a set of users of the first Web service, and a set of users of the second Web service Storage means for storing the second computed recognition tag and the third computed recognition tag in association with a third recognition tag for uniquely identifying the user in the union A server device is provided.

In another embodiment of the present invention, a server device for providing a Web service, the first identification tag for uniquely identifying the first user among the users of the Web service A first computed recognition tag that has been subjected to an irreversible computation, and a second user of the web service that has a relationship with the first user in the web service among the users of the web service The server apparatus is characterized by having a transmission means for transmitting in a state where the second calculated recognition tag obtained by performing the irreversible calculation on the second recognition tag is specified.

In yet another embodiment of the present invention, a first computed recognition obtained by performing an irreversible computation on a first recognition tag for uniquely identifying a first user among web service users. The irreversible operation is performed on the vote and the second identification tag for uniquely identifying the second user having a relationship with the first user in the Web service. There is provided a server device comprising receiving means for receiving the two calculated recognition tags from the Web service server.

In still another embodiment of the present invention, the first identification tag for uniquely identifying the first user who uses the first Web service among the users of the first Web service is calculated. The first web service use that has been made and the second user who uses the first web service that has a relationship in the first web service with the first user, Receiving means for receiving, from a first server that provides the first Web service, a second calculated recognition tag obtained by performing the above calculation on a second recognition tag for uniquely identifying a person; A third computed identification tag obtained by computing a third identification tag for uniquely identifying the first user using the second web service among the users of the second web service; A relationship between the first user and the second Web service. A fourth computed recognition tag obtained by performing the above computation on a fourth identification tag for uniquely identifying a third user who uses the second web service among the second web service users; And a receiving unit that receives the second Web service from a second server that provides the second Web service.

In still another embodiment of the present invention, the first identification tag for uniquely identifying the first user who uses the first Web service among the users of the first Web service is the first one. The first calculation result obtained by performing the calculation and the second user who uses the first Web service having a relationship in the first Web service with the first user are designated as the first Web service. A relationship between the second recognition tag obtained by performing the first calculation on the second recognition tag for uniquely identifying among the users, and the second user and the first Web service. A third is obtained by performing the first calculation on a third identification tag for uniquely identifying a third user who uses the first Web service, among the users of the first Web service. Computed recognition tag and a first familiarity coefficient between the first user and the second user Receiving means for receiving a second familiarity factor between the second user and the third user from a first server providing the first Web service; and A computing means for performing a second computation on the intimacy coefficient and the second intimacy coefficient to calculate a third intimacy coefficient; and the third intimacy coefficient as a second server of the second Web service. There is provided a server device characterized by having a transmission means for transmitting to the server.

In still another embodiment of the present invention, the first identification tag for uniquely identifying the first user who uses the first Web service among the users of the first Web service. A first calculation result obtained by performing the first calculation and a second user who uses the first Web service having a relationship in the first Web service with the first user. From a first server providing the first Web service, a second computed recognition tag obtained by performing the first calculation on a second recognition tag for uniquely identifying among users The first calculation is included in a third identification tag for uniquely identifying the receiving means for receiving and the second user using the second Web service among the users of the second Web service. To the third operation result, and the second user and the second Web service. The first calculation is applied to the fourth identification tag for uniquely identifying the third user who uses the second Web service having the relationship among the second Web service users. Receiving means for receiving a fourth computed recognition tag and additional information of the third user in the second Web service from a second server providing the second Web service; Calculating means for calculating a third intimacy coefficient by applying a second operation to the intimacy coefficient of 1 and the second intimacy coefficient;
There is provided a server device comprising: a transmission unit configured to transmit the additional information and the third intimacy coefficient to the first server.

In still another embodiment of the present invention, the first identification tag for uniquely identifying the first user who uses the first Web service among the users of the first Web service. A first calculation result obtained by performing the first calculation and a second user who uses the first Web service having a relationship in the first Web service with the first user. A second computed identification tag obtained by performing the first computation on a second identification tag for uniquely identifying among users, and between the first user and the second user Receiving means for receiving the first familiarity coefficient from the first server that provides the first Web service, and the second user using the second Web service as the second Web service. A third identification tag for performing the first calculation on the third identification tag to uniquely identify the user The calculation result and the third user who uses the second Web service having a relationship in the second Web service with the second user are uniquely identified among the second Web service users. For this purpose, a fourth calculated recognition tag obtained by performing the first calculation on the fourth recognition tag, additional information in the second Web service of the third user, and the second user Receiving means for receiving a second familiarity coefficient between the second user and the third user from a second server that provides the second Web service, and the first familiarity coefficient and the second user A server device comprising: an arithmetic means for performing a second operation on the intimacy coefficient of the first to calculate a third intimacy coefficient; and a transmitting means for transmitting the additional information to the first server. Provided.

In still another embodiment of the present invention, the first identification tag for uniquely identifying the first user who uses the first Web service among the users of the first Web service is calculated. The first calculation result applied and the second user who uses the first Web service having a relationship in the first Web service with the first user are assigned to the first Web service user. A second identification card that has been subjected to the above-mentioned calculation on a second identification tag for uniquely identifying the first identification web, and the first web that has a relationship in the first web service with the second user A third calculated identification tag obtained by performing the calculation on a third identification tag for uniquely identifying a third user who uses the service among the users of the first Web service; Receiving means for receiving from a first server providing a first Web service; A fourth calculation result obtained by calculating a fourth identification tag for uniquely identifying the first user using the second Web service among the users of the second Web service; A fifth for uniquely identifying a fourth user who uses the second Web service having a relationship with one user in the second Web service among the second Web service users A fifth calculated recognition tag obtained by performing the calculation on the identification tag, and a third user who uses the second Web service having a relationship in the second Web service with the fourth user, A second Web service that provides the second Web service with a sixth calculated recognition tag obtained by performing the calculation on a sixth identification tag for uniquely specifying among the users of the second Web service; Receiving means for receiving from the server, That server device is provided.

According to the present invention, it is possible to integrate and handle attribute information and acquaintance relationships of a plurality of Web services without disclosing information held by each Web service to the outside. As a result, it is possible to provide a method for realizing an acquaintance candidate recommendation service that is broader and more accurate than in the past.

The presently considered modes for carrying out the present invention will be described below. Since the scope of the present invention is clearly defined by the appended claims, the present invention should not be construed in a limited sense to these forms. These forms are described merely for the purpose of illustrating the general principles of the invention.

In order to make the expression simple and clear, a user of a certain Web service or a certain network service may be called a node. One node can be expressed by one identification tag in the Web service or network service. The relationship between nodes is called an edge. For example, in the Web service X, when expressing an event that the user A and the user B are acquaintances, “the node represented by IDxa and the node represented by IDxb are connected by an edge”. Express. Further, in a certain Web service or a certain network service, the relationship between nodes composed of nodes and edges is referred to as a social graph of the Web service or network service.

(Embodiment 1)
FIG. 2 shows an example of an embodiment of the present invention, a network service N server 200 that provides search and prioritization of acquaintance candidates beyond the Web service to the Web service, a Web service X server 210, and a Web service A Y server 220 and a Web service Z server 230 are connected via a network 291. In addition, the Web service X server 210, the Web service Y server 220, and the Web service Z server 230 are connected to the information terminal 251 of the user A, the information terminal 252 of the user F, the information terminal 253 of the user E, and the network. 292 is connected. The connection relationship shown in FIG. 2 is an example. Although only the information terminals of user A, user F, and user E necessary for explanation are displayed in FIG. 2, the server 210 of the web service X, the server 220 of the web service Y, and the web service Z The server 230 is connected to information terminals of all users of each Web service through some network. In FIG. 2, the network 291 and the network 292 may be the same network or different networks.

FIG. 3 is an example of a schematic configuration diagram of the server 200 of the network service N. The server of the network service N includes transmission / reception means 310, storage means 320, calculation means 330, and search means 340. The storage unit 320 includes node information 321, edge information 322, a calculated recognition tag table 323, a reception information table group 325, a closeness table 326, a web service weight table 327, and a closeness coefficient-added graph table 328. The transmission / reception means 310 is connected to a network and can communicate.

FIG. 4 is an example of a schematic configuration diagram of a Web service X server. The server 210 of the Web service X includes a transmission / reception unit 410, a storage unit 420, a calculation unit 430, an HTML generation unit 440, a search unit 450, and a closeness coefficient generation unit 460. In addition, the storage unit 420 includes node information 421, edge information 422, and a calculated recognition tag table 423 in the Web service X.

FIG. 5 is an example of a schematic configuration diagram of a Web service Y server. The server 220 of the Web service Y includes a transmission / reception unit 510, a storage unit 520, a calculation unit 530, an HTML generation unit 540, a search unit 550, and a closeness coefficient generation unit 560. In addition, the storage unit 520 includes node information 521, edge information 522, and a calculated recognition tag table 523 in the Web service Y.

FIG. 6 is an example of a schematic configuration diagram of a Web service Z server. The server 230 of the Web service Z includes transmission / reception means 610, storage means 620, calculation means 630, HTML generation means 640, search means 650, and closeness coefficient generation means 660. The storage unit 620 includes node information 621, edge information 622, and a calculated recognition tag table 623 in the Web service Z.

Now, it is assumed that the Web service X has the social graph shown in FIG. This indicates that, for example, user A and user B, user B and user E, and user F are acquaintances in Web service X. Here, “acquaintances in the Web service X” can be interpreted as meaning that the server of the Web service X is stored as having some relationship between recognition tags. For example, a pair of identification tags is stored in a database. In FIG. 7, for example, the fact that user A and user B are acquaintances is that the node 701 represented by the identification tag IDxa and the node 702 represented by the identification tag IDxb are joined by an edge 703. Expressed in shape. In FIG. 7, the relationship between other users is the same. In addition, the same expression is used when expressing the relationship between users thereafter.

The social graph of the web service X shown in FIG. 7 is stored in the storage unit 420 of the server of the web service X shown in FIG. In the node information 421 of the storage unit 420, identification tags IDxa, IDxb, IDxe, and IDxf are stored in association with information held by the user represented by each identification tag in the Web service X. Further, as shown in FIG. 7, the edge information 422 stores information that IDxa and IDxb, IDxb and IDxe, and IDxb and IDxf are respectively connected by edges.

Further, the node information 421 stores the node 701 of the user A represented by IDxa and the additional information Pxa751 and Qxa752 in association with each other. Here, the additional information Pxa is information on the user A sent to the server 200 of the network service N in steps described later. On the other hand, Qxa is information of user A that is not sent to the server 200 of the network service N and is disclosed only to the user of the Web service X. Note that “disclosed only to the user of the web service X” may mean that the web service X can be acquired from the server of the web service X when the user is using the web service X. Hereinafter, the additional information of the user T in the Web service S that is sent to the server 200 of the network service N and may be disclosed to users other than the Web service S is referred to as shared additional information, and is denoted as Pst. To do. Further, the additional information of the user T in the Web service S that is disclosed to only the user of the Web service S without being sent to the server 200 of the network service N is referred to as dedicated additional information and is expressed as Qst. Here, two types of information, shared additional information and dedicated additional information, are collectively referred to as additional information. Further, additional information Pxb and Qxb are stored in association with the node of user B represented by IDxb. As shown in FIG. 7, user E has additional information Pxe and Qxe, and user F has additional information Pxf and Qxf. There may or may not be shared additional information associated with each node.

The information stored in the edge information 422 is obtained by connecting the nodes represented by IDxa and IDxb, similarly between IDxb and IDxe, and similarly between IDxb and IDxf at the edges. Any format can be used as long as it contains information that it is present. As an example of the expression format of such information, FOAF (Friend Of A Friend) is known (see, for example, Non-Patent Document 5). This is XML (eXtended Markup Language) / RDF (Resource Description Language). This is an expression format for expressing the relationship between nodes and edges. For example, a situation in which the node ID xa and the node ID xb are connected by an edge is expressed in the form shown in FIG. Here, as an example, the FOAF representation format is shown, but any representation format may be used as long as it represents the social graph shown in FIG. The same applies to FIGS. 9 and 10 described later.

Further, it is assumed that the Web service Y has a social graph shown in FIG. This represents, for example, that the user A and the user C, the user C and the user E, and the user C and the user F are acquaintances in the web service Y.

The social graph of the web service Y shown in FIG. 9 is stored in the storage unit 520 of the server of the web service Y shown in FIG. In the node information 521 included in the storage unit 520, identification tags IDya, IDyc, IDye, and IDyf are stored in association with the information held in the web service Y of each user. Further, the edge information 522 includes, as indicated by a one-dot chain line in FIG. 9, between a node represented by IDya and a node represented by IDyc, similarly between IDyc and IDye, and similarly between IDxc and IDyf. The information that each is connected by an edge is stored. The information stored in the edge information 522 includes a node between IDya and a node represented by IDyc, a connection between IDyc and IDye, and a connection between IDyc and IDyf. As long as the information is included, any format may be used. As shown in FIG. 9, the node information 521 includes additional information Pya and Qya associated with the node IDya, additional information Pyc and Qyc associated with the node IDyc, additional information Pye and Qye associated with the node IDye, Additional information Pyf and Qyf are stored in association with the node IDyf. There may or may not be shared additional information associated with each node.

Further, it is assumed that the Web service Z has a social graph shown in FIG. This represents, for example, that the user A and the user D, the user D and the user E, and the user D and the user G are acquaintances in the Web service Z.

The social graph of the web service Z illustrated in FIG. 10 is stored in the storage unit 620 of the server of the web service Z illustrated in FIG. In the node information 621 included in the storage unit 620, identification tags IDza, IDzc, IDze, and IDzg are stored in association with information held in the Web service Z by the user represented by each identification tag. Further, as indicated by the broken line in FIG. 10, the edge information 622 includes a node between IDza and a node represented by IDzc, similarly between IDzc and IDze, and similarly between IDzc and IDzg. Information connected at the edge is stored. In the information stored in the edge information 622, between the node represented by IDza and the node represented by IDzc, similarly between IDzc and IDze, and similarly between IDzc and IDzg, are coupled by edges. As long as the information is included, any format may be used. Further, as shown in FIG. 10, the node information 621 includes additional information Pza and Qza associated with the node IDza, additional information Pzd and Qzd associated with the node IDzd, additional information Pze and Qze associated with the node IDze, Additional information Pzg and Qzg are stored in association with the node IDzg. There may or may not be shared additional information associated with each node.

Next, an example of processing in which the Web service X server 210 registers information in the network service N server 200 will be described with reference to the flowchart shown in FIG.

First, the calculation means 430 of the Web service X server performs an irreversible calculation on the user identification card stored in the node information 421 of the storage means, or is stored only in the storage means 420 of the Web service X server. The encryption operation is performed by the key to be used. The node information of the Web service X stores an identification tag of each node represented by a black dot in FIG. The calculation means 430 generates the calculation results F (IDxa), F (IDxb), F (IDxe), and F (IDxf) of these recognition tags (step S1101). F represents an irreversible operation or an encryption operation using a key stored only in the storage unit 420 of the server of the Web service X.

Any irreversible computation and encryption computation algorithm may be used here. As an example of an irreversible calculation algorithm, there is a so-called hash calculation such as MD-5 (Message Digest Algorithm 5) or SHA1 (Secure Hash Algorithm 1). Examples of algorithms for encryption operations include DES (Data Encryption Standard), so-called symmetric encryption, triple DES, and AES (Advanced Encryption Standard). Another example of encryption is RSA (Rivest Shamir Adleman) called so-called asymmetric encryption. The secret key used for the symmetric encryption method and the private key (secret key) used for the decryption in the asymmetric encryption method are not sent to the server 200 of the network service N, but stored in the storage means 420 of the server of the Web service X. Preferably only is stored.

Next, the calculation unit 430 of the server of the Web service X converts the social graph shown in FIG. 7 into a graph using the calculation result. That is, the information shown in FIG. 7 is converted into information that F (IDxa) and F (IDxb), F (IDxb) and F (IDxe), and F (IDxb) and F (IDxf) have a relationship. This is stored in the computed recognition tag table 423 in the storage means 420 (step S1102). The expression format used when expressing the relationship between the calculation results may be anything as long as this information is included.

FIG. 12 shows an example of the calculated recognition tag table in the Web service X. “Identification slips for each user in Web service X” 1210 and each “calculated recognition slip” 1220 are stored in association with each other. Further, each calculated recognition tag and “the calculation result of the recognition tag combined at the edge by the Web service X” 1230 are stored in association with each other. In FIG. 12, for example, the identification tag IDxb 1212 and the calculation result F (IDxb) 1222 are associated with each other, and F (IDxa) 1234, F (IDxe) 1232, and F (IDxf) 1233 are associated with each other. Further, shared additional information 1250 and dedicated additional information 1260 are associated with each of the recognition tags of the Web service X.

As shown in FIGS. 7, 9, and 10, the user A uses the three services of the web service X, the web service Y, and the web service Z. Here, the user A next transmits the identification tag IDya and IDza of the user A in the service other than the web service X used by the user A to the server 210 of the web service X from the information terminal 251. To do. Next, the transmission / reception means 410 of the server of the Web service X receives these identification tags IDya and IDza (step S1103). Next, the calculation means of the server of the Web service X performs the calculation on these identification tags received from the information terminal of the user A, and obtains F (IDya) and F (IDza) as a result. Here, the server of the Web service X calculates the calculation results F (IDya) and F (IDza) received from the information terminal other than the Web service X of the user A from the information terminal of the user A. Another method for the Web service X 210 to obtain the above two calculation results is that the user A logs in to the Web service Y through the information terminal, and then the server 220 of the Web service Y generates the calculation result F (IDya). In addition, this may be transmitted to the server 210 of the Web service X and received by the server 210 of the Web service X. Similarly, the user A logs in to the Web service Z through the information terminal, the server 230 of the Web service Z generates the calculation result F (IDza), and transmits it to the server 210 of the Web service X. May be received by the server 210 of the Web service X.

Next, these F (IDya) and F (IDza) are stored in the calculated recognition tag table 423 in association with the calculation result F (IDxa) of the identification tag IDxa of the user A who is the transmission source (step S1104). Referring to FIG. 12, the calculation results F (IDya) 1241 and F (IDza) 1242 obtained here are displayed in a column 1240 of “calculation results of recognition tags other than Web service X” in the calculated recognition tag table 423. F (IDxa) 1221 is stored in association with it.

Next, the transmission / reception means 410 of the server of the Web service X includes the table of the portion excluding the column 1210 of the “service X identification tag” and the column 1260 of “dedicated additional information” in the calculated recognition tag table 423. It transmits to the server 200 of the network service N (step S1105). Therefore, it is difficult or impossible for the server of the network service N to know which user of the Web service X each node is. That is, the server of the network service N has received an anonymous social graph from the server of the Web service X.

Next, processing in the server 200 of the network service N that has received the information in the table from the server of the Web service X will be described with reference to FIG. The transmission / reception means 310 of the server of the network service N receives the table transmitted from the server of the Web service X, and stores it as one of the reception information table group 325 (step S1301). Next, the received information is stored in the calculated recognition tag table 323 of the server of the network service N in association with the identification tag for uniquely identifying the node in the network service N (step S1303).

FIG. 14 shows an example of the calculated recognition tag table 323 after execution of this step. First, a recognition tag in the network service N corresponding to each “calculated recognition tag” included in the column 1220 of the received calculated recognition tag is generated. For example, an identification tag IDna 1411 of the service N corresponding to the received F (IDxa) is generated. IDna may be anything as long as it can uniquely identify this node in the network service N. The same applies to the other identification tags in the network service N. In the calculated recognition tag table 323 in the server of the network service N, the information in the column of “the calculation result of the recognition tag other than the Web service X” 1240 among the received information is the column of the “received calculated recognition tag”. 1420. For example, IDna 1411 includes F (IDxa) 1421 and F (IDya) 1241 and F (IDza) 1242 sent from the server of Web service X as the “calculation result of identification tag other than Web service X” “reception”. Are stored as F (IDya) 1422 and F (IDza) 1423 in the column 1420 of “computed recognition card”. Further, the information of the column of “calculation result of recognition tag combined with edge in Web service X” 1230 received from the server of Web service X is “calculated recognition tag combined with edge in service N” column 1430. Is remembered. For example, the “computed recognition tag combined with F (IDxa) and edge in Web service X” F (IDxb) 1231 sent from the server of Web service X is “computed already combined with edge in service N” It is stored as “recognition tag” F (IDxb) 1431.

FIG. 15 is a graph display of the information of FIG. 14 for easy understanding. This graph includes the node 1511 represented by the identification tag IDna1411 of the network service N, which is associated with the calculated identification tag F (IDxa), F (IDya), and F (IDza) received in step S1301. A node 1512 represented by an identification tag IDnb1412 associated with the completed identification tag F (IDxb), a node 1513 represented by an identification tag IDne1413 associated with the computed recognition tag F (IDxe), and a computed recognition tag F It consists of a node 1514 represented by an identification tag IDnf1414 associated with (IDxf). These pieces of information are included in the column 1410 of the “service N identification tag” and the column 1420 of the “computed identification tag received” in FIG. Further, the graph of FIG. 15 shows a node 1511 represented by IDna and a node 1512 represented by IDnb, a node 1512 represented by IDnb and a node 1513 represented by IDne, and a node 1512 represented by IDnb. And a node 1514 represented by IDnf are connected by an edge. The above is equivalent to the information stored in the column 1430 of “calculated recognition slips combined at the edge in service N” in FIG. Each node in FIG. 15 has additional information. This is equivalent to the information stored in the “shared additional information” column 1440 of FIG. The schema of the computed recognition tag table 323 in FIG. 14 is merely an example, and the expression format is not limited as long as the graph information in FIG. 15 is stored.

Each node represented by IDna, IDnb, IDne, and IDnf stored in the calculated identification tag table 323 of the network service N is the user A, the user B, the user E, and the user in the server 210 of the Web service X. Correspond to each person F. However, each identification tag is received by the server 200 of the network service N in an irreversible operation or an encryption operation in a state where it cannot be decrypted except by the server of the network service N. Therefore, the server of the network service N cannot guide to which user each node in the calculated table 323 corresponds. That is, the above process means that the social graph in the Web service X is mapped to the graph of the anonymous node on the network service N.

Next, a process for mapping the social graph in the Web service Y to the graph of the anonymous node on the network service N will be described. Each step of the process in the Web service Y server is the same as the process in the Web service X server shown in the flowchart of FIG.

First, the computing unit 530 of the Web service Y server performs an irreversible computation on the user identification card stored in the node information 521 of the storage unit, or is stored only in the storage unit 520 of the Web service Y server. The encryption operation is performed by the key to be used. The node information of the Web service Y stores an identification tag of each node represented by a black dot in FIG. The calculating means 530 generates the calculation results F (IDya), F (IDyc), F (IDye), and F (IDyf) of these recognition tags.

Any irreversible operation and encryption operation algorithm may be used here, but the same algorithm as the irreversible operation or encryption operation used in step S1101 in the server of the Web service X.

Next, the computing unit 530 of the server of the Web service Y converts the social graph shown in FIG. 9 into a graph of a computed recognition tag. That is, the information shown in FIG. 9 is combined at the edge between F (IDya) and F (IDyc), between F (IDyc) and F (IDye), and F (IDyc) and F (IDyf). Is stored in the calculated recognition tag table 523 in the storage means 520. The expression format used for expressing the relationship between the calculation results may be anything as long as the above information is included.

FIG. 16 shows an example of the calculated recognition tag table 523 in the Web service Y. “Identification slips for each user in Web service Y” 1610 and each “calculated recognition slip” 1620 are stored in association with each other. Further, each calculated recognition tag is stored in association with “calculation result of recognition tag combined with edge in Web service Y” 1630. In FIG. 16, for example, an identification tag IDyf 1614 and its calculation result F (IDyf) 1624 are associated with each other, and F (IDyc) 1634 coupled with an edge is associated with each other. Further, shared additional information 1650 and dedicated additional information 1660 are associated with each of the identification tags of the Web service Y.

As shown in FIGS. 7 and 9, the user F uses two services of the web service X and the web service Y. Next, the user F transmits the identification tag IDxf of the user F in the service other than the web service Y used by the user F from the information terminal 252 to the server 220 of the web service Y. Next, the transmission / reception means 510 of the server of Web service Y receives this identification tag IDxf. Next, the calculation means of the server of the Web service Y performs the calculation on the identification tag received from the information terminal of the user F, and obtains the result F (IDxf). Here, the server of the web service Y calculates the calculation result F (IDxf) after receiving the identification tag IDxf other than the web service Y of the user F from the information terminal of the user F. As another method for the Web service Y 220 to obtain the above calculation result, the user F logs in to the Web service X through the information terminal, and the server 210 of the Web service X generates the calculation result F (IDxf). In addition, this may be sent to the server 220 of the Web service Y and received by the server 220 of the Web service Y. In the description of the calculation F and the user F, the same F is used, but can be distinguished from the context.

Next, this F (IDxf) is stored in the calculated recognition tag table 523 in association with the calculation result F (IDyf) of the identification tag IDyf of the user F who is the transmission source. Referring to FIG. 16, the calculation result F (IDxf) 1644 obtained here is stored in the column 1640 of “calculation result of recognition tag other than Web service Y” in the calculated recognition tag table 523 and F (IDyf) 1624. Saved in association.

Next, the transmission / reception means 510 of the server of the Web service Y removes the portion of the calculated recognition tag table 523 excluding the column 1610 of the “service Y identification tag” and the column 1660 of the “dedicated additional information” from the network service N. To the server 200. Therefore, it is difficult or impossible for the server of the network service N to know which user of the Web service Y each node is. That is, the server of the network service N has received an anonymous social graph from the server of the web service Y.

Next, processing in the server 200 of the network service N that has received the above table from the server of the Web service Y will be described with reference to the flowchart of FIG. The transmission / reception means 310 of the server of the network service N receives the information of the table sent from the server of the Web service Y, and stores this as one of the reception information table group 325 (step S1301). Next, the table information received from the server of the Web service Y is stored in the calculated recognition tag table 323 of the server of the network service N by the following procedure. This information is previously received from the Web service X and stored in addition to the information already stored in the calculated recognition tag table 323 as shown in FIG.

The procedure for adding the table information received from the Web service Y server to the information already received from the Web service X server shown in FIG. 14 will be described with reference to FIGS. As a result, FIG. 14 shows the calculated recognition tag table 323 before addition, and FIG. A part of FIG. 16 is information received from the server of the Web service Y. First, the search means 340 of the network service N uses the calculation result F (IDxf) 1644 of the recognition tag other than the Web service Y received from the server of the Web service Y as a search key. The line 1420 of the “computed recognition tag received” is searched. Then, F (IDxf) 1424 in the fourth row is obtained as a search result. Next, F (IDyf) 1624 associated with F (IDxf) 1644 in FIG. 16 is added as F (IDyf) 1725 on the fourth line of “received computed recognition tag” in the computed recognition tag table 323. This is because the user having IDxf in the Web service X and the user having IDyf in the Web service Y are the same user. This corresponds to the storage in association (step S1304). However, since an irreversible operation or an encryption operation that cannot be decrypted by the server of the network service N is performed, the server of the network service N cannot derive IDxf and IDyf. For this reason, it is impossible to know that the node represented by IDnf is the server of the network service N and represents the same user F in the web service X and the web service Y.

Next, the search unit 340 searches for the same recognition tag in the “received calculated recognition tag” 1420 of the calculated recognition tag table 323 of the network service N after executing step S1304 (step S1305). In this case, the same identification tag is not searched (step S1306).

Next, the search means 330 of the network service N searches the “calculated recognition tag” 1620 received from the server of the Web service Y for items that are not in the “calculated recognition tag received” 1420. Here, as shown in FIG. 14, there is no calculated recognition tag F (IDyc) 1635 and F (IDye) 1634 in the column of “received calculated recognition tag” 1420 in the calculated recognition tag table 323. Therefore, a new line is added to the computed recognition form table 323, a new recognition form in the network service N called IDnc1715 is generated, and F (IDyc) 1726 is stored in association with this recognition form IDnc1715. Also, a new identification tag in the network service N called IDnh1716 is generated, and F (IDye) 1727 is stored in association with this identification tag IDnh. This means that a new node having F (IDyc) represented by the identification tag IDnc and F (IDye) represented by IDnh has been generated (step S1308).

Next, among the calculation results in the row 1630 of the “calculation result of the identification tag combined at the edge of the Web service Y” received by the search unit 340 from the server of the Web service Y, the calculated recognition tag table of the network service N An unregistered calculation result is searched in a column 1430 of “calculated recognition tag combined at the edge in service N” 323. Next, this unregistered calculation result is stored in the column 1430 of the “calculated recognition tag combined at the edge in service N” of the calculated recognition tag table 323 of the network service N and received from the server of the Web service Y. Each identification tag is stored in association with each other (step S1309). Specifically, referring to FIG. 17, F (IDyc) 1731 is associated with F (IDya) 1422, F (IDyc) 1734 is associated with F (IDyf) 1725, and F (IDyc) 1726 is associated with F (IDyc) 1726. (IDye) 1735, F (IDyf) 1736, and F (IDya) 1738 are stored in association with F (IDye) 1727, and F (IDyc) 1737 is stored. As a result, the edge of the social graph of the Web service Y shown in FIG. 9 is superimposed on the graph of the calculated recognition tag on the network service N shown in FIG.

Next, the information received from the server of Web service Y is stored in the “shared additional information” column 1440 of the calculated recognition tag table 323 of the network service N in association with the calculated recognition tag that has received the shared additional information. (Step S1310). Specifically, Pya 1742 is associated with F (IDya) 1422, Pyf 1744 is associated with F (IDyf), Pyc 1745 is associated with F (IDyc) 1726, and Pye 1746 is associated with F (IDye) 1727. .

Thus, the information received from the Web service Y server is added to the information previously received from the Web service X server shown in FIG. 14, and the state shown in FIG. 17 is obtained. FIG. 18 is a graph display of FIG. 17 for simplicity. In FIG. 18, a calculated recognition tag with a number starting with 14 is information before being received from the Web service Y in FIG. 14. Further, the calculated recognition tag with a number beginning with 16 or 17 is information received from the server of the Web service Y and added to the calculated recognition tag table 323. Each number corresponds to FIG. 15, FIG. 16, and FIG. In FIG. 18, the solid line represents an edge on the network service N caused by the edge of the Web service X. Similarly, a dashed line represents an edge on the network service N caused by the edge of the Web service Y.

FIG. 18 shows a result of generating a composite graph of two graphs limited in each of the web service X and the web service Y shown in FIGS. 7 and 9 in the server of the network service N. However, it is not a graph of recognition tags for each of the X and Y Web services, but a composite graph with each calculated recognition tag as a node. Therefore, in the server of the network service N, it is difficult or impossible to specify which user in each of the web services X and Y each node is.

Next, a process for mapping a social graph in the Web service Z to a graph of an anonymous node on the network service N will be described. Each step of the processing in the Web service Z server is the same as the processing in the Web service X server shown in the flowchart of FIG.

First, the computing means 630 of the Web service Z server performs an irreversible computation on the user identification card stored in the node information 621 of the storage means, or is stored only in the storage means 620 of the Web service Z server. The encryption operation is performed by the key to be used. The node information of the Web service Y stores an identification tag of each node represented by a black dot in FIG. The calculation means 530 generates calculation results F (IDza), F (IDzd), F (IDze), and F (IDzg) of these recognition tags.

Any irreversible operation and encryption operation algorithm may be used here, but the same algorithm as the irreversible operation or encryption operation used in step S1101 in the server of the Web service X.

Next, the computing unit 630 of the server of the Web service Z converts the social graph shown in FIG. 10 into a graph of a computed recognition tag. That is, the information shown in FIG. 10 is represented by edges between F (IDza) and F (IDzd), between F (IDzd) and F (IDze), and between F (IDzd) and F (IDzg). The information is converted into information that is combined, and this is stored in the computed recognition form table 623 in the storage unit 620. The expression format used for expressing the relationship between the calculation results may be anything as long as the above information is included.

FIG. 19 shows an example of the calculated recognition tag table 623 in the Web service Z. The “service Z identification tag” 1910 and each “calculated identification tag” 1920 are stored in association with each other. Further, each calculated recognition tag and “the calculation result of the recognition tag combined at the edge by the Web service Z” 1930 are stored in association with each other. In FIG. 19, for example, an identification tag IDze 1913 and its calculation result F (IDze) 1923 are associated with each other, and F (IDyd) 1933 coupled with an edge is associated with each other. Further, shared additional information 1950 and dedicated additional information 1960 are associated with each of the recognition tags of the Web service Z.

As shown in FIGS. 7, 9, and 10, the user E uses three services of the web service X, the web service Y, and the web service Z. Next, the user E transmits the identification tag IDxe and IDye of the user E in the service other than the web service Z used by the user E from the information terminal 253 to the server 230 of the web service Z. Next, the transmission / reception means 610 of the server of the Web service Z receives these identification tags IDxe and IDye. Next, the calculation means 630 of the server of the Web service Z performs the calculation on these recognition tags received from the information terminal of the user E, and obtains F (IDxe) and F (IDye) as a result. Here, the server of the Web service Z calculates the calculation results F (IDxe) and F (IDye) after receiving the identification card IDxe and IDye other than the Web service Z of the user E from the information terminal of the user F. did. As another method for the Web service Z 230 to obtain the above calculation result, the user E logs in to the Web service X through the information terminal, and the server 210 of the Web service X generates the calculation result F (IDxe). In addition, this may be sent to the server 230 of the Web service Z and received by the server 230 of the Web service Z. Similarly, the user E logs in to the web service Y through the information terminal, the server 220 of the web service Y generates the calculation result F (IDye), and sends it to the server 230 of the web service Z. The server 230 of the Web service Z may receive it.

Next, these F (IDxe) and F (IDye) are stored in the calculated recognition tag table 623 in association with the calculation result F (IDze) of the identification tag IDze of the user E who is the transmission source. Referring to FIG. 19, the calculation results F (IDxe) 1943 and F (IDye) 1944 obtained here are associated with F (IDze) 1923 in the column “1940” of “the calculation result of the recognition tag other than the Web service Z”. Saved.

Next, the transmission / reception means 610 of the server of the Web service Z transmits a portion of the calculated recognition tag table 623 excluding the “service Z recognition tag” column 1910 to the server 200 of the network service N. For this reason, it is difficult or impossible for the server of the network service N to know which user of the Web service Z each node is. That is, the server of the network service N has received an anonymous social graph from the server of the web service Z.

Next, the processing in the server 200 of the network service N that has received the table from the server of the Web service Z will be described again with reference to the flowchart of FIG. The transmission / reception means 310 of the server of the network service N receives the table sent from the server of the Web service Z and stores it as one of the reception information table group 325 (step S1301). Next, the table information received from the server of the Web service Z is stored in the calculated recognition tag table 323 of the server of the network service N according to the following procedure. This information is previously received from the Web service X and the Web service Y, and stored in addition to the information already stored in the calculated recognition tag table 323 as shown in FIG.

The procedure for adding the table information received from the Web service Z server to the information already received from the Web service X and Web service Y servers shown in FIG. 17 will be described with reference to FIGS. . As a result, FIG. 17 shows the calculated recognition tag table 323 before the addition and FIG. 20 after the addition. A part of FIG. 19 is information received from the server of the Web service Z. First, the search means 340 of the network service N uses the calculation results F (IDxe) 1943 and F (IDye) 1944 of recognition tags other than the Web service Z from the server of the Web service Z as search keys, and the calculated recognition tag of FIG. A column 1420 of “received computed recognition tag” in the table 323 is searched. Then, F (IDxe) 1723 in the third row and F (IDye) 1727 in the sixth row are obtained as search results. Next, F (IDze) 1923 associated with F (IDxe) 1943 in FIG. Store in the same location as (IDxe) 1723. This is because the user having IDx in the Web service X and the user having IDze in the Web service Z are the same user, so that the identification tag in these two different Web services is the identification tag IDne of the network service N. It is equivalent to memorizing in association with. Similarly, F (IDze) 1923 associated with F (IDye) 1944 in FIG. 19 is changed to the sixth row of column 1720 of “Received computed recognition tag” in the computed recognition tag table 323 in FIG. Store in the same location as (IDye) 1723. This is because the user who has IDye in the Web service X and the user who has IDze in the Web service Z are the same user, so that the identification tag in these two different Web services is the identification tag IDnh of the network service N. (Step 1304). However, since an irreversible operation or an encryption operation that cannot be decrypted by the network service N server is performed, it is difficult or impossible for the network service N server to obtain IDxe, IDye, and IDze. . For this reason, it is difficult or impossible to know that the node represented by IDne represents the same user E in the web service X, the web service Y, and the web service Z in the server of the network service N. .

Next, the search unit 340 searches for the same recognition tag in the “received calculated recognition tag” 1420 of the calculated recognition tag table 323 of the network service N after executing step S1304 (step S1305). Then, a match between F (IDze) associated with IDne 1713 in step S1304 and F (IDze) associated with IDnh 1716 in step S1304 is detected (step S1306). This means that in the network service N, the node represented by IDne and the node represented by IDnh are the same user. Therefore, the IDne line and the IDnh line are integrated in the calculated recognition tag table 323 of the network service N. As shown in FIG. 20, the IDne 1713 is integrated here. As a result, F (IDxe) 2022, F (IDye) 2023, and F (IDze) 2024 are associated with each other and stored in the row 1420 of the “received computed recognition tag” in association with the identification tag IDne1713. . In addition, for “calculated recognition tag combined at the edge in service N” 1430, F (IDyc) 1737 in FIG. 17 is stored in association with IDne as F (IDyc) 2034 in FIG. Are integrated (step S1307).

Next, the search means 330 of the network service N searches the “computed recognition form” 1920 received from the server of the Web service Z for items that are not in the “computed recognition form received” 1420. Here, as shown in FIG. 17, there is no calculated recognition tag F (IDzd) 1922 and F (IDzg) 1924 in the column of “received calculated recognition tag” 1420 of the calculated recognition tag table 323. Therefore, a new line is added to the calculated recognition tag table 323 to generate a new identification tag in the network service N called IDnd2016, and F (IDzd) 2026 is stored in association with this recognition tag IDnd. Further, a new identification tag in the network service N called IDng 2017 is generated, and F (IDzg) 2027 is stored in association with this identification tag IDnh. This means that two new nodes having F (IDzd) represented by the identification tag IDnd and F (IDzg) represented by IDng have been generated (step S1308).

Next, among the calculation results in the row 1930 of the “calculation result of the identification tag combined with the edge in the Web service Z” received from the server of the Web service Z by the search unit 340, the calculated recognition tag table of the network service N An unregistered calculation result is searched in a column 1430 of “calculated recognition tag combined at the edge in service N” 323. Next, this unregistered calculation result is displayed in the column 1430 of the “calculated recognition tag combined at the edge in service N” of the calculated recognition tag table 323 of the network service N and received from the server of the Web service Z. Each identification card is stored in association with each other (step S1309). Specifically, referring to FIG. 20, F (IDzd) 2031 is associated with F (IDza) 1721, F (IDzd) 2033 is associated with F (IDze) 2024, and F (IDzd) 2026 is associated with F (IDzd) 2026. (IDza) 2036, F (IDze) 2037, and F (IDzg) 2038 are stored in association with F (IDzg) 2027, and F (IDzd) 2039 is stored. As a result, the edge of the social graph of the Web service Z shown in FIG. 10 is superimposed on the graph of the calculated recognition tag on the network service N shown in FIG.

Next, the information received from the server of the Web service Z is stored in the “shared additional information” column 1440 of the calculated recognition tag table 323 of the network service N in association with the calculated recognition tag that has received the shared additional information. (Step S1310). Specifically, Pya 2041 is associated with F (IDza) 1721, Pze 2043 is associated with F (IDze) 2024, Pzd 2046 is associated with F (IDzd) 2026, and Pye 2047 is associated with F (IDzg) 2027. To do.

Thus, the information received from the Web service Z server is added to the information previously received from the Web service X and Y servers shown in FIG. 17, and the state shown in FIG. 20 is obtained. FIG. 21 is a graphical representation of FIG. 20 for simplicity. In FIG. 21, a calculated recognition tag with a number starting with 14 is information received from the Web service X in FIG. 14. In addition, a computed recognition tag with a number starting with 16 or 17 is information received from the server of the Web service Y. Further, a computed recognition tag with a number beginning with 19 or 20 is information received from the Web service Z server. Each number corresponds to FIG. 15, FIG. 16, FIG. 17, FIG. In FIG. 21, the solid line represents the edge of the Web service X, the alternate long and short dash line represents the edge of the Web service Y, and the broken line represents the edge of the Web service Z.

By the method described so far, the server 200 of the network service N receives the graph of the calculated identification tag from each of the server 210 of the web service X, the server 220 of the web service Y, and the server 230 of the web service Z. This was superimposed. As a result, the server of the network service N inherits the characteristics of the social graph of each Web service shown in FIG. 21 and does not know which user on which Web service each node is. Graphs can be generated.

The computed recognition tag table in FIG. 20 is one expression format of the graph shown in FIG. However, in the method according to the present invention, the method for expressing the anonymous graph in the network service N may be any method as long as the graph of FIG. 21 can be expressed. Next, three embodiments of the present invention will be described that are possible using this graph. The following three embodiments are realized using the graph shown in FIG. Therefore, the above description is common to the following three embodiments.

(Embodiment 2)
In this embodiment, a method of using the information of the graph shown in FIG. 21 to allow a user who uses a plurality of Web services to collectively manage acquaintances in different Web services on one Web service. provide. In the method according to the present invention, this is possible without the network service N obtaining an identification tag or social graph of each Web service user. A method for realizing this will be described below with reference to the flowchart of FIG.

As shown in FIGS. 7, 9, and 10, user A uses Web services X, Y, and Z. Now, the user A transmits a request for acquiring acquaintance information in another Web service on the Web service X to the server of the Web service X through the information terminal 251 (step S2201). The transmission / reception means 410 of the server of the Web service X receives this. Next, the transmission / reception means 410 of the server of the Web service X associates a request for acquiring acquaintance information in another Web service with the calculated recognition tag F (IDxa) of the user A to the server 200 of the network service N. Transmit (step S2202). This is received by the transmission / reception means 310 of the server of the network service N. Next, in the following procedure, a node in the network service N that is coupled with the node having the received F (IDxa) at the edge is searched (step S2203). Specifically, the “received computed recognition tag” column 1420 of the computed recognition tag table 323 shown in FIG. 20 is searched using the received F (IDxa) as a search key. Then, F (IDxa) 1421 is obtained as a search result. This is a node represented by IDna in the network service N. Next, the search means 340 searches for the “computed recognition tag combined at the edge in service N” 1430 associated with IDna. Then, F (IDxb), F (IDyc), and F (IDzd) are obtained as search results. Based on these, the “received computed recognition tag” column 1420 of the computed recognition tag table 323 of the network service N is searched again. Then, F (IDxb) is a calculated recognition tag received from the Web service X, and is a node represented by IDnb in the network service N. Similarly, F (IDyc) is a node represented by IDnc received from the Web service Y, and F (IDzd) is a node represented by IDnd received from the Web service Z (step S2204).

The request received by the server 200 of the network service N in step S2201 is a request for information of an acquaintance other than the Web service X. Therefore, the transmission / reception means 210 associates the recognition tag IDnc with the shared additional information Pyc, and associates the recognition tag IDnd with the shared additional information Pzd, and transmits them to the server 210 of the Web service X (step S2205). Next, it transmits to the server of Web service X. The server 410 of the Web service X receives these.

Next, the HTML generation unit 440 of the server of the Web service X generates information (HTML code) using, for example, HTML (HyperText Mark-up Language) as information including the additional information Pyc and the additional information Pzd, and is transmitted and received. The means 410 transmits this to the information terminal 251 of the user A (step S2206). The information terminal 251 of the user A receives and analyzes this HTML code and displays it on the display means.

FIG. 23 is an example of a screen on which the HTML code in step S2206 is displayed on the display means of the information terminal 251 of the user A. A Web service X screen 2300 is displayed in the Web browser window. This screen has a display 2301 indicating that it is the screen of user A and a list 2302 of acquaintances. One of the features of an embodiment of the present invention is that an acquaintance other than the Web service X can be displayed in the acquaintance list. In this screen, the acquaintance 2320 and the acquaintance 2330 correspond to the node represented by IDnc and the node represented by IDnd on the network service N, respectively. In this example, the additional information Pyc is a nickname 2321, an avatar 2322, and a web service name 2323, and the additional information Pzd is a nickname 2331, an avatar 2332, and a web service name 2333.

Here, the effect of this embodiment concerning this invention is demonstrated. According to the present embodiment, even if the server of the network N does not obtain the identification slips and social graphs of the users of the web services X, Y, and Z, for example, acquaintances exceeding the web service as shown in FIG. Can be viewed. Further, according to the present embodiment, the server 210 of the Web service X can display information about users of other services as shown in FIG. 23, for example, without obtaining the identification slips of the users of the Web services Y and Z. . In general, from a business viewpoint, there is a strong demand that each Web service does not want to disclose user information and social graphs to other Web service providers. According to the present embodiment, each Web service can use information obtained by the union of social graphs of all Web services while satisfying this requirement.

(Embodiment 3)
Embodiment 3 of the present application is a method that can recommend a person who is not a direct acquaintance as a highly acquainted candidate using an anonymous graph that spans multiple Web services as shown in FIG. provide. According to the present embodiment, this is possible without the network service N obtaining an identification tag or social graph of each Web service user. A method for realizing this will be described below with reference to the flowchart of FIG.

Now, the user A transmits a request for information for displaying a web page or the like to the server of the web service X through the information terminal 251 (step S2401). The server 210 of the Web service X receives this. Next, the transmission / reception means 410 of the server of the Web service X transmits a request for recommendation of an acquaintance candidate to the server 200 of the network service N in association with the user A's calculated recognition tag F (IDxa) (step S100). S2402). Then, in the server 200 of the network service N that has received this, the node IDna having F (IDxa) and, for example, a node having 2 as a distance in the anonymous graph of the network service N from the calculated recognition vote table 323 according to the following procedure Search for. Here, a node having a distance of 2 refers to another node in which a node coupled by an edge is further coupled by another edge. That is, the distance between nodes connected by one edge is 1.

Referring to FIG. 21, it is apparent that nodes having a node IDna and a distance of 2 are nodes IDne, IDnf, and IDng. Hereinafter, an example of a method for deriving this will be described with reference to FIG. First, the search means 340 of the network service N searches for the “computed recognition tag combined at the edge in the service N” 1430 associated with the received F (IDxa) (step S2403). . As a result, F (IDxb), F (IDyc), and F (IDzd) are obtained. These correspond to the nodes represented by IDna having F (IDxa) and the three nodes IDnb, IDnc, and IDnd joined at the edge (step S2404). Next, the calculation means searches the calculated recognition tag table 323 to search for a node that is connected to the three nodes by an edge. As a result, F (IDxe), F (IDxf), F (IDxa), F (IDye), F (IDyf), F (IDya), F (IDza), F (IDze), and F (IDzg) are obtained. It is done. These correspond to the nodes IDna, IDne, IDnf, and IDng. That is, it is derived that there are three nodes IDne, IDnf, and IDng representing indirect acquaintances having a distance of 2 on the anonymous graph of node IDna and network service N (step S2406). In the anonymous graph of the network service N, if a node having IDna and 2 as a distance can be searched, the method may not be the method described above.

Next, the search means of the network service N calculates the number of nodes existing between each of the node IDne, IDnf, and IDng and the node IDna in the anonymous graph. Referring to FIG. 21, it is clear that there are three nodes IDnb, IDnc, and IDnd between node IDne and node IDna. Similarly, it can be seen that there are two nodes between the node IDnf and the node IDna, and one node between the node IDng and the node na (step S2407). In the following, an example of a method for deriving this will be described with reference to FIG. For example, in order to check the number of nodes between the node IDne and the node IDna, the search means 340 of the network service N performs a “computed recognition tag combined at the edge of the service N at the edge between the IDna2011 row and the IDne1713 row. ”Column 1430 is searched for a common identification tag. As a result, three of F (IDxb), F (IDyc), and F (IDzd) are obtained. This corresponds to the node IDnb, the node IDnc, and the node IDnd. In a similar manner, two nodes, node IDnb and node IDnc, are obtained as nodes between node IDnf and node IDna. In a similar manner, one of IDnds is obtained as a node between the node IDng and the node IDna. In the anonymous graph of the network service N, if the number of nodes between IDna and a node having a distance of 2 can be derived, the method may not be the method described above.

Based on the node having a distance of 2 on the network service N and the number of nodes between each node and the node IDna obtained in step S2407, the computing unit 330 stores the storage unit 320 in FIG. A closeness table 326 as shown in FIG. Referring to FIG. 25A, a closeness table, for example, the identification tag IDne2511 is associated with the number of intermediate nodes 2521 between IDna. IDne is associated with a computed recognition tag F (IDxe) 2531 possessed by this node. This is because the retrieval means 340 has received the computed recognition tag received from the Web service X from the “received computed recognition tag” 1420 and “shared additional information” 1440 in the IDne row of the computed recognition tag table 323 shown in FIG. It can be realized by searching for shared additional information. However, the node IDng having a distance of 2 from IDna is an anonymous node representing the user G who does not use the Web service X. For this reason, there is no calculated recognition tag for the Web service X associated therewith. Therefore, it is preferable that the node ID ng is not included in the response to the acquaintance candidate recommendation request from the server 210 of the Web service X in step S2402. This is because the user G is not a user of the Web service X, so even if the number of intermediate nodes related to the node ID ng is sent to the server 210 of the Web service X, the server of the Web service X sends information about the user G. This is because the user G cannot be displayed in response to the request from the information terminal 251 of the user A.

Nodes having a distance of 2 on an anonymous graph in which three social graphs of Web services X, Y, and Z are superimposed represent acquaintances of acquaintances, that is, indirect acquaintances on some web services. That is, the number of intermediate nodes 2520 corresponds to the number of acquaintances common to all Web services between indirect acquaintances. Even with acquaintances of acquaintances, it is presumed that users with many common acquaintances have a closer relationship. This is the reason why the table including the number of intermediate nodes as shown in FIG.

Next, the transmission / reception means 310 of the server of the network service N sends the closeness table shown in FIG. 25A to the server 210 of the Web service X (step S2408). Next, the server 210 of the Web service X receives this. The search means 450 of the server of the Web service X uses the received “calculated recognition tag of Web service X” 2530 as a search key, and reads the “calculated recognition tag” row 1220 of the calculated recognition tag table 423 shown in FIG. Search for. As a result, identification tags IDxe and IDxf of service X corresponding to F (IDxe) and F (IDxf) are obtained. This is an identification tag for Web service X of user E and user F, respectively, and the received common node numbers 3 and 2 are associated (step S2409). This means that there are three common acquaintances between user A and user E, and two common acquaintances between user A and user F.

Next, the HTML generation unit of the server of the Web service X generates an HTML code including dedicated additional information Qxe 1263 and Qxf 1264 of the user E and the user F as the acquaintance candidate of the user A. At this time, in accordance with the number of common nodes, the user E generates an HTML code that is displayed as an acquaintance candidate that is more likely to become a more direct acquaintance than the user F. Next, the transmission / reception means 410 of the Web server X transmits this HTML code to the information terminal 251 of the user A (step S2410). The information terminal of user A receives the HTML code, analyzes it, and displays it on the display means (step S2411).

FIG. 26 is an example of a screen on which the HTML code in step S2411 is displayed on the display unit of the information terminal 251 of the user A. A web service screen 2600 is displayed in the web browser window. This screen has a display 2601 indicating that it is the screen of user A, and a list 2602 of acquaintance candidates with priorities. In this example, the user E2610 and the user F2620 are displayed in the list of acquaintance candidates together with dedicated additional information and priority. The priority order 2613 is displayed in association with the dedicated additional information 2611 and 2612 of the user E, and the priority order 2623 is displayed in association with the dedicated additional information 2621 and 2622 of the user F. In the example of FIG. 26, the name and face photo of each user are used as the dedicated additional information. By this display, the user A who has the name “Jack White” has more common acquaintances than the user F who has the name “Susan Berry”, and is more likely to be a direct acquaintance. You can know that you are a high-priority user. The priority order display of the acquaintance candidates may not be a character string display as shown in FIG. For example, the priority order may be expressed by a method of displaying with a graphic such as an icon or the display order.

Here, the effect of this embodiment will be described. In the example shown in this embodiment, the users A, E, and F are all users of the Web service X. Therefore, even with the conventional method, it is possible in the prior art that the server of the Web service X can guide that the user A and the users E and F are indirect acquaintances of nodes having a distance of 2. there were. Therefore, the user X server can display the user E and the user F as acquaintance candidates on the information terminal of the user A. However, looking at the social graph possessed by the web service X shown in FIG. 7, in the conventional method, if another web service is included between the user A and the user E in the server of the web service X, three people are included. It is difficult or impossible to detect that there is a common acquaintance. Similarly, in the Web service X server, if another Web service is included between the user A and the user F, it is difficult or impossible to detect that there are two common acquaintances. . Therefore, with the conventional method, it is difficult or impossible to realize an acquaintance candidate recommendation service as shown in FIG.

However, according to this embodiment, the server of the network service N does not obtain the user's identification tag or social graph of the web service X, and the server of the web service X obtains the social graph of another web service. Therefore, it is possible to prioritize acquaintance candidates beyond one Web service as shown in FIG. Independence between Web services is required mainly from a business viewpoint, and according to the present embodiment, there is an effect that both independence of Web services and more accurate acquaintance candidate recommendation can be achieved.

In the above description according to the present embodiment, the server of the network service N transmits the information of the closeness table 326 shown in FIG. 25A to the server of the Web service X, and the Web service X received in step S2409 The server transmits the dedicated additional information of the Web service X to the information terminal of the user A. For this reason, the closeness table shown in FIG. 26A does not include information related to the node of IDng that is not the user of the Web service X. In another example of this embodiment, the information of the familiarity table as shown in FIG. 26B may be transmitted to the Web service X server. In this case, the closeness table includes the node ID ng of the user who is not using the Web service X. Furthermore, the shared additional information 2570 obtained from the calculated recognition tag table of the network service N is associated with each recognition tag of “node having distance 2 from IDna” 2550. In this case, when the server 210 of the Web service X that has received the familiarity table generates an HTML code to be sent to the information terminal of the user A, the shared additional information is displayed together with the priority order. This makes it possible to provide a service that recommends users other than the Web service X as candidates for acquaintance in addition to the user of the Web service X. However, there is generally less shared additional information than dedicated additional information. For example, even if the web service X dedicated additional information Pxe or Pxf includes various information such as names, self-introductions, diaries, photos, and videos, the shared additional information disclosed to other web service users is a nickname. And often avatars. Therefore, in this example using the familiarity table of FIG. 26B, a service that recommends a wider range of users as acquaintance candidates is realized, but information obtained by the user regarding the acquaintance candidates tends to be less.

(Embodiment 4)
This embodiment realizes ranking of acquaintances including acquaintances on other Web services by associating coefficients representing the intimacy of the relationship between users with each edge of the anonymous graph in the network service N. Or recommend more acquaintance candidates.

In general, the strength and familiarity of acquaintances between users in each Web service are not the same. For example, acquaintances using voice calls are more likely to be more intimate than anonymous acquaintances who met on SNS. In addition, even among users in the same Web service, users who frequently exchange messages are more likely to be intimate than users who only browse each other's written information. In the second embodiment, when displaying users of other Web services in the Web service X, closeness with the users is not considered. In the third embodiment, when a user who has a distance of 2 on the integrated anonymous graph in the network service N is recommended as an acquaintance candidate, prioritization is performed in consideration of closeness. The density index is the number of acquaintances in common with acquaintance candidates. That is, in the second and third embodiments, the assumption is made that both the acquaintance relationship in each Web service and network service N, that is, the edges of the graph represent all equal intimacy.

On the other hand, the fourth embodiment uses the premise that the closeness of the edges constituting the social graph of each Web service is different. In other words, in the integrated anonymous graph in the network service N, information related to prioritization when the web service displays an acquaintance list or a list of acquaintance candidates after storing an affinity coefficient associated with each edge. Then, the network service N server 200 transmits the information to each Web service server.

In this embodiment, a closeness coefficient is added to the computed recognition tag tables 423, 523, and 623 of each Web service server. In the above description, FIG. 12, FIG. 16, and FIG. 19, which represent each, have no description of the familiarity coefficient. Therefore, in the following, how the familiarity coefficient is processed in the calculated recognition tag table of each Web service server will be described using the calculated recognition tag table of the Web service X server as an example.

FIG. 27 shows a closeness coefficient in the social graph of the Web service X. In the example shown in this embodiment, the edge of the social graph is a directional edge. For this reason, there are two familiarity coefficients μxab 2711 and μxba 2712 at the edge between the user A node 701 represented by the identification tag IDxa and the user B node 702 represented by IDxb. Similarly, μxbe and μxeb are at the edge between IDxb and IDxe, and μxbf and μxfb are between IDxb and IDxf, respectively. In the method according to the present embodiment, the social graph of each Web service may be an edge with direction or an edge without direction. In the case of a directional edge as shown in the present embodiment, there are two familiarity coefficients per edge, but in the case of an edge without direction, there is one familiarity coefficient per edge. In the following, the familiarity coefficient for the directional edge from the node T to the node U in the Web service S is expressed as μstu. In the example of this embodiment, 0 ≦ μstu ≦ 1 is set for simplicity.

In the server 210 of the Web service X, the calculated recognition tag table 423 is generated from the social graph shown in FIG. 27 according to the steps S1101 and S1102. FIG. 28 shows a calculation result of “recognition tag of service X” 1210, “recognition tag of calculation” 1220, and “recognition tag combined at the edge by Web service X” in the calculated recognition tag table including the familiarity coefficient. "1230 is extracted and shown. However, in the present embodiment, a closeness coefficient is associated with each of the calculated recognition tags in the “recognition tags combined at the edge of the Web service X” and stored. For example, the edge with direction 2721 having the node ID xa shown in FIG. 27 as the start point and the node ID xb as the end point is set to F (IDxb) 2831 associated with F (IDxa) 2821 in the calculated recognition tag table shown in FIG. Further, it is expressed by μxab 2711 further related thereto. Similarly, a directional edge 2722 having the node IDxb shown in FIG. 27 as the start point and the node IDxa as the end point is F (IDxa) 2832 associated with F (IDxb) 2822 in the computed recognition tag table shown in FIG. Further, it is expressed by μxba2712 further associated therewith. The other directional edges shown in FIG. 27 are also stored in the calculated recognition tag table shown in FIG.

The intimacy coefficient may be any numerical value indicating the strength of connection between nodes in the social graph of the Web service X. For example, if the Web service X is an SNS, the storage unit 420 stores the number of times the past user A has browsed the information of the user B. Next, the familiarity coefficient generation means 460 stores this number of times as the familiarity coefficient μxab in the calculated recognition tag table shown in FIG. On the contrary, the number of times that the user B browses the information of the user A is defined as an intimacy coefficient μxba of the edge from the user B node to the user A node. If the Web service X provides a voice call or chat service, the number of calls and chats and the length of time can be used as the intimacy coefficient. In addition, each user inputs information indicating the strength of the human relationship with other users from the input means of each information terminal, and then transmits this information to the Web service X server. The information may be a closeness coefficient. For example, the strength of the relationship with the user B felt by the user A is input to the information terminal 251 as a numerical value from 1 to 5. The information terminal 251 of the user A transmits the numerical value to the server 210 of the Web service X. The server of the Web service X uses the numerical value as the familiarity coefficient μxab. In addition, intimacy coefficients need not be associated with all edges. In addition, some or all of the familiarity coefficients of the social graph in the Web service X may be the same value.

Next, as described in steps S <b> 1103 to S <b> 1105 in FIG. 11, a part of the information in the calculated recognition tag table of the Web service X is transmitted to the server 200 of the network service N.

The Web service Y server 220 and the Web service Z server 230 also create a computed identification tag table including a closeness coefficient by the same method, and transmit a part thereof to the server 200 of the network service N. The processing in the Web service Y server 220 and the Web service Z server 230 is the same as the processing already described for the Web service X server in the example of the present embodiment, and a description thereof will be omitted.

The server 200 of the network service N includes a calculation result of the recognition tag combined with an edge in the calculated recognition tag table including the closeness coefficient from each of the servers of the Web services X, Y, and Z. The closeness coefficient associated with each, the calculation result of the recognition tag of another Web service, and the shared additional information are received.

The server 200 of the network service N receives the information from the Web services X, Y, and Z in this way, and creates the calculated recognition tag table 323 of the network service N in the procedure as shown in FIG. In this embodiment, the closeness coefficient received from the server of each Web service is associated with each of the calculated recognition slips “computed recognition slips combined at the edge in service N” 1430 of the calculated recognition slip table 323. And memorized. For the sake of explanation, FIG. 29 shows an excerpt of the “calculated recognition tag combined at the edge in the service N” in the calculated recognition tag table 323 in the present embodiment. For example, in the node IDna 2911 of the network service N, the familiarity coefficient μxab 2932 is stored in association with the calculated recognition tag F (IDxb) 2931. Referring to FIG. 28, the F (IDxb) 2831 and the μxab 2711 associated therewith in the Web service X server are received by the network service N server 200 and stored in the calculated recognition tag table 323. It has been done. Similarly, referring to FIG. 29 again, F (IDyc) 2933 and μyac 2934 received from the server of Web service Y are stored in association with each other. Further, F (IDzd) 2935 and μzad 2936 received from the Web service Z server are stored in association with each other.

Next, the calculation unit 330 of the server of the network service N uses the Web service weight table 327 stored in the storage unit and the calculated identification tag table 323 including the closeness coefficient shown in FIG. Create The procedure will be described below.

First, the Web service weight table will be described. The web service weight coefficient represents the contribution of each web service regarding the familiarity. For example, assume that the Web service X is a voice call service used by people who already know their real names, the Web service Y is an SNS service, and the Web service Z is a blog service used by anonymous people. In this case, it is presumed that the acquaintances of the web service Y are more familiar than the acquaintances of the web service Z, and the acquaintances of the web service X are more acquainted than the acquaintances of the web service Y. The Web service weight table expresses such a situation in the form of a weight coefficient for each Web service.

FIG. 30 is an example of a web service weight table in an example of the present embodiment. The web service weight table has a web service identification tag 3010 and a weight coefficient 3020. In FIG. 30, the web service X identification tag http: // www. X. The weighting coefficient λx3021 is stored in association with com (3011). Similarly, weight coefficients λy 3022 and λz 3023 are stored in association with the identification tags of the Web services Y and Z, respectively.

Next, the computing unit 330 generates a graph table 328 with a closeness coefficient from the computed recognition tag table 323 and the Web service weight table 327. FIG. 31 is an example of a graph table with a closeness coefficient in the present embodiment. The familiarity coefficient-added graph table stores each node 1210 of the network service N in association with the recognition tag of each node connected by an edge and the familiarity coefficient κ. In FIG. 31, for example, the closeness coefficients κab, κac, and κad in the network service N are stored in association with the node ID nb 3121, the node ID nc 3123, and the node ID nd 3125 that are coupled to the node ID na at the edge in the network service N, respectively. The

For example, the closeness coefficient κab 3122 in the network service N is obtained by κab = μxab · λx / (λx + λy + λz). Similarly, κac3124 and κad3126 are obtained by κac = μyac · λy / (λx + λy + λz) and κad = μzad · λz / (λx + λy + λz), respectively. The intimacy coefficient μstu received from each Web service in FIG. 29 to the intimacy coefficient κtu in the network service N in FIG. 31 uses λs · μstu as the Web service weighting factor λs in FIG. It is expressed as a value normalized by the sum of the service weight coefficients. For example, it is assumed that Web service X is a voice call service used by people who know their real names, and Web service Z is a blog service used by anonymous people. Further, for example, it is assumed that the strength of the acquaintance relationship in these Web services is expressed by the familiarity coefficient λx: λz = 2: 1. Then, all the intimacy coefficients μxtu sent from the Web service X server 210 are twice the weight of all the intimacy coefficients μztu sent from the Web service Z server 230, and the intimacy coefficients of the network service N Contributes to κtu. For example, if the closeness coefficients of the edges of the web services X, Y, and Z are all the same at 1, the closeness coefficient κtu of the network service N is the weighting coefficient of the web service in which the edge exists λs. Then, λs / (λx + λy + λz) is obtained.

In other words, the closeness coefficient κtu of each edge of the anonymous graph in the network service N has a property that becomes larger as the acquaintance relationship between users in each Web service is stronger and the more important the Web service is. Yes. The calculation method of the closeness coefficient of each edge of the anonymous graph in the network service N described in the present embodiment is an example, and any method can be used as long as the closeness coefficient calculation method has such a property. Good.

In the description of the present embodiment, the example shown in FIG. 21 is used for simplicity as an example of the anonymous graph in the network service N. As a more general example, there may be a case where two anonymous nodes of the network service N are connected by an edge in a plurality of Web services. An example is shown in FIG.

In FIG. 32, black dots represent nodes in the network service N. A solid line is an edge in the Web service X, a dashed line is an edge in the Web service Y, and a dotted line is an edge in the Web service Z. For example, in network service N, user A represented by node ID na3201 represents a situation in which user B represented by node ID nb3202 in Web service X knows by directional edge 3211. The same applies to other directional edges. That is, in FIG. 32, the six arrows between the node IDna and the no IDnb indicate that the users A and B are acquaintances with each other in all of the Web services X, Y, and Z. Similarly, four arrows between the node ID nb3202 and the node ID nc 3204 indicate that the user B and the user C are acquainted with each other in the web services Y and Z. Further, three arrows between the node ID nb3202 and the node ID nd 3203 indicate that the user B and the user D are acquainted with each other in the web service X, and the user B knows the user D in the web service Y. , User D does not know user B.

At this time, it is assumed that the closeness coefficient is associated with the edge of the social graph of each Web service. That is, in the server 210 of the Web service X, μxab is associated with the edge 3211, μxba is associated with the edge 3212, μxbd is associated with the edge 3231, and μxdb is associated with the edge 3232. In the Web service Y server 220, μyab is associated with the edge 3213, μyba with the edge 3214, μybc with the edge 3221, μycb with the edge 3222, and μybd with the edge 3233, respectively. Further, in the server of the Web service Z, μzab is associated with the edge 3215, μzba with the edge 3216, μzbc with the edge 3223, and μzcb with the edge 3224, respectively.

At this time, for example, the familiarity coefficient κtu of the directed edge from the node T to the node U in the network service N can be calculated as follows, thereby realizing the property that the familiarity coefficient of the network service N should satisfy.

In the example of FIG. 32, for example, the familiarity coefficient κab of the directional edge from the node IDna to IDnb in the network service N is (λx · μxab + λy · μyab + λz · μzab) / (λx + λy + λz). Similarly, the familiarity coefficient κba of the directional edge from the node IDnb to the node IDna is (λx · μxba + λy · μyba + λz · μzba) / (λx + λy + λz). In the case of an edge between the node IDnb and the node IDnd in FIG. 32, the closeness coefficient κbd of the directional edge from the node IDnb to the node IDnd is (λx · μxbd + λy · μybd) / (λx + λy + λz). Conversely, the familiarity κdb of the edge with the direction from the node IDnd to the node IDnb is λx · μxdb / (λx + λy + λz). Note that the calculation method shown here is an example, and any calculation method may be used as long as the calculation method realizes the property that the closeness coefficient of the network service N should satisfy.

Next, the server 200 of the network service N transmits information included in the graph with closeness 328 to each Web service X server, and each Web service server uses this information to improve the function of the Web service. Two examples of the method according to this embodiment will be described with reference to FIG. The processing of the first example described below is the same as that of the second embodiment, and the processing of the second example is almost the same as that of the third embodiment. Therefore, the first example will refer to the flowchart shown in FIG. 22 and the second example will refer to the flowchart shown in FIG.

In the first example of the present embodiment, when the server of the Web service X displays a list including acquaintances in other services, as shown in FIG. Display possible acquaintances with priority. For example, in the display of the acquaintance list 2302 in FIG. 23, acquaintances considered to have a stronger acquaintance relationship can be displayed with high priority. The outline of the process will be described below with reference to the flowchart of FIG.

In the first example of the present embodiment, first, as shown in steps S2201 and S2202 in FIG. 22, the user A of the web service X transmits an acquaintance information acquisition request to the server 210 of the web service X, and further the web service The server of X makes a search request to the server 200 of the network service N using the calculated recognition tag F (IDxa) as a search key. Next, the network service N server searches the graph table 328 with a closeness coefficient using the identification tag IDna of the network service N associated with F (IDxa) as a search key. Referring to FIG. 31, IDnb, IDnc, and IDnd, and the closeness coefficients κab, κac, and κad associated with each of them are obtained as search results (corresponding to step S2204). These and the shared additional information Pyc and Pzd are sent to the server 210 of the Web service X (corresponding to step S2205). Then, the server 210 of the Web service X generates an HTML code including the shared additional information Pxb or dedicated additional information Qxb of the user B, the shared additional information Pyc of the user C, and the shared additional information Pzd of the user D, It transmits to the information terminal of user A (corresponding to step S2206). Here, in the present embodiment, the computing means 430 of the server of the Web service X uses the associated services in descending order of the closeness coefficients κab, κac, κad of the network service N received from the server 200 of the network service N. HTML code is displayed to indicate that the user's priority is high. The information terminal of user A receives the HTML code, analyzes it, and displays an image as shown in FIG. 23 on the display means (corresponding to step S2207). In the present embodiment, the screen of FIG. 23 is displayed so that the order of the familiarity coefficients received from the server of the network service N can be understood. Indication of intimacy rank performed here is any method that can express intimacy order, such as enumeration in descending order of intimacy coefficient, intimacy coefficient display, or intimacy order display. Various forms may be used. In the first example of the present embodiment described individually, the additional information in the information terminal of the user A is obtained by using the closeness coefficient of the directional edge from the user A to the users B, C, and D. Prioritizing the display. In another example, prioritization of additional information display in the information terminal of the user A may be performed based on the closeness coefficients of the directed edges in the reverse direction, that is, κba, κca, and κda. Further, in another example, the addition in the information terminal of the user A based on the result of evaluating the bidirectional closeness coefficient by taking the average of the closeness coefficients of the directed edges in the opposite directions. Prioritization of information display may be performed.

In the second example of the present embodiment, when the server of the Web service X displays an acquaintance of an acquaintance as an acquaintance candidate as shown in FIG. 26 on the information terminal of the user, the server of the network service N Using the received closeness coefficient, acquaintance candidates that are considered to be more intimate are preferentially displayed. The outline of the processing will be described below with reference to the flowchart of FIG.

In the second example of the present embodiment, first, as shown in steps S2401 and S2402 of FIG. 24, the server 210 of the Web service X is triggered by the Web page display request from the user A, and the server 200 of the network service N is triggered. Send an acquaintance candidate acquisition request to Next, the network service N server searches the graph table 328 with a closeness coefficient using the identification tag IDna of the network service N associated with F (IDxa) as a search key (corresponding to step S2403). Referring to FIG. 31, IDnb, IDnc, and IDnd and κab, κac, and κad associated with each of them are obtained as search results (corresponding to step S2404). Next, using IDnb, IDnc, and IDnd as search keys, the node connected with the edge in the familiarity-added graph table 328 and the closeness coefficient of the edge having the node as the end point are searched again (corresponding to step S2405). Referring to FIG. 31, IDne, IDnf, and IDng nodes are obtained as search results. Furthermore, as a result of the search, three familiarity coefficients κbe, κce, and κde having the node IDne as an end point, two familiarity coefficients κbf and κcf having the node IDnf as an end point, and an intimacy coefficient having an IDng as an end point κdg is obtained (corresponding to step S2406). FIG. 33 obtained as a result of this search is a graph display of the recognition tag in the network service N and the closeness coefficient for each directional edge.

Next, for each of the nodes IDne, IDnf, and IDng at a secondary distance from the node IDna, the computing unit 330 calculates the sum of the closeness coefficients κ of the edges between the node IDna and A density table 326 is created (corresponding to step S2407). Referring to FIG. 33, for example, the sum of the familiarity coefficients for IDnf is κab + κac + κcf + κbf. Similarly, the sum of the familiarity coefficients for IDne is κab + κac + κad + κbe + κce + κde. Similarly, the sum of the familiarity coefficients for IDng is κad + κdg. An example of the familiarity table 326 in this embodiment created from this is shown in FIG.

Next, the identification card, the shared additional information, and the sum of the closeness coefficients in the network service N of these three nodes are associated and transmitted to the server of the Web service X (corresponding to step S2408). Next, the server of the Web service X generates an HTML code including the received sum of the familiarity coefficients or its rank, and transmits it to the information terminal 251 of the user A (corresponding to step S2410). Next, the information terminal of user A analyzes the received HTML code, and displays acquaintance candidates with priorities on the display means (corresponding to step S2411).

As a result, an example of the screen displayed on the display means of the information terminal 251 of the user A is as shown in FIG. 26 described in the second embodiment. In the case of Embodiment 2, the priority between the user A and the acquaintance candidate displayed here was evaluated by the number of common friends. In this embodiment, the priority order between the user A and the acquaintance candidate displayed here is also evaluated by the sum of the intimacy coefficients of the edges between the user A and the acquaintance candidate in the network service N.

Here, the effect of the present invention according to the present embodiment will be described. The first example shown in the present embodiment has the following effects in addition to the effects shown in the first embodiment. In the first embodiment, information on acquaintances in other Web services can be displayed while maintaining the independence of each Web service. In the method according to the present embodiment, it is possible to evaluate the strength of the relationship with the acquaintance and preferentially display the acquaintance with a strong relationship.

In addition, the second example shown in the present embodiment has the following effects in addition to the effects shown in the second embodiment. In the third embodiment, while maintaining the independence of each Web service, a user who is at a distance 2 in the union of anonymous social graphs of all Web services can be prioritized and recommended as an acquaintance candidate. It was. In the third embodiment, a common number of people is used for the prioritization, but in this embodiment, the sum of the familiarity coefficients of edges in the union of anonymous social graphs is used. As a result, it is possible to recommend an acquaintance candidate that more accurately reflects the strength of the human relationship in the actual human society.

It is a figure showing an example of the social graph which a some web service has. It is a figure showing an example of the system configuration which enforces this invention. It is a schematic block diagram of the server of the network service N which implements this invention. 2 is a schematic configuration diagram of a server of Web service X. FIG. 2 is a schematic configuration diagram of a server of a Web service Y. FIG. 2 is a schematic configuration diagram of a server of a Web service Z. FIG. It is a figure showing the social graph which Web service X has. It is an example figure for demonstrating the social graph expression format by FOAF. It is a figure showing the social graph which Web service Y has. It is a figure showing the social graph which Web service Z has. It is a flowchart of an example of the process in the server of Web service X. It is a figure showing an example of the node information in the server of Web service X, and edge information. It is a flowchart of an example of the process in the server of the network service N by the method of this invention. It is a figure showing an example of the calculated recognition tag table of the network service N which memorize | stored the information received from Web service X. FIG. 5 is a graph display of a computed recognition tag table of the network service N that stores information received from the Web service X. FIG. It is a figure showing an example of the node information and edge information in the server of Web service Y. It is a figure showing an example of the calculated recognition tag table of the network service N which memorize | stored the information received from Web service X and Y. FIG. 6 is a graph display of a computed recognition tag table of a network service N that stores information received from Web services X and Y. FIG. It is a figure showing an example of the node information in the server of Web service Z, and edge information. It is a figure showing an example of the calculated recognition tag table of the network service N which memorize | stored the information received from Web service X, Y, and Z. FIG. 5 is a graph display of a calculated recognition tag table of a network service N that stores information received from Web services X, Y, and Z. FIG. It is a flowchart showing the process for displaying an acquaintance other than Web service X on the information terminal of the user of Web service X. It is a figure showing an example of the screen which displayed the acquaintance other than Web service X on the information terminal of the user of Web service X. It is a flowchart showing the process for assigning a priority order and recommending an acquaintance candidate to the information terminal of the user of Web service X. It is a figure showing the example of the closeness table in the server of Web service N. FIG. 10 is an example of a screen for recommending acquaintance candidates with priorities assigned to information terminals of users of a Web service X. It is a figure showing an example of the closeness coefficient in the social graph of Web service X. It is a figure showing an example of the calculated recognition tag table with the closeness coefficient of Web service X. It is a figure showing a part of calculated recognition tag table with the closeness coefficient of the network service N. It is an example figure showing the web service weight table of network service N. It is a figure showing an example of the graph table with closeness of the network service N. It is a conceptual diagram showing that the edge in a some web service is shared between nodes. It is the figure which displayed the closeness coefficient in the network service N in the graph. It is an example figure of the closeness table in the network service N.

Explanation of symbols

200 Server Device 210 Web Service X Server 220 Web Service Y Server 230 Web Service Z Server 251 User A Information Terminal 252 User F Information Terminal 253 User E Information Terminal 291 Network 292 Network

Claims (19)

The first server of the first Web service performs an operation on the first identification tag for uniquely identifying the user who uses the first Web service among the users of the first Web service. The first calculated recognition tag and the second identification tag for uniquely identifying the user using the second Web service among the users of the second Web service. Receiving means for receiving, from the first server, a second computed recognition tag that has been subjected to the computation at the server;
Receiving means for receiving, from the second server, a third computed recognition form obtained by performing the computation on the second recognition form by the second server of the second Web service;
Using the third computed recognition form as a search key, the computed recognition form received from the first server and the second server is retrieved, and the second computed recognition form and the third computation are searched. A search means for detecting a match of a completed identification tag;
In association with a third identification tag for uniquely identifying the user in the union of the set of users of the first Web service and the set of users of the second Web service, Storage means for storing two calculated recognition cards and the third calculated recognition tag;
The server apparatus characterized by having. The server apparatus according to claim 1, wherein the calculation has irreversibility. The operation uses symmetric encryption processing,
A secret key for decrypting the first computed recognition tag and the second computed recognition tag is stored in the first server;
The server apparatus according to claim 1, wherein a secret key for decrypting the third computed recognition tag is stored in the second server. The operation uses asymmetric cryptography,
A secret key for decrypting the first computed recognition tag and the second computed recognition tag is stored in the first server;
2. The server apparatus according to claim 1, wherein a secret key for decrypting a result obtained by performing the calculation on the third calculated identification tag is stored in the second server. A server device for providing a Web service,
A first computed recognition tag obtained by performing an irreversible operation on the first recognition tag for uniquely identifying the first user among the users of the Web service, and the first user A second operation obtained by performing the irreversible operation on a second identification tag for uniquely identifying the second user of the Web service having a relationship in the Web service among the users of the Web service A server device comprising: a transmission unit configured to transmit a completed identification tag in an associated state. A first computed recognition tag obtained by performing an irreversible operation on the first recognition tag for uniquely identifying the first user among Web service users, the first user, and the first user A second computed recognition tag obtained by performing the irreversible computation on a second recognition tag for uniquely identifying a second user having a relationship in the Web service among the users is provided as the Web service. A server device comprising receiving means for receiving from the server. A first computed recognition tag obtained by computing the first identification tag for uniquely identifying a first user using the first web service among the users of the first web service; A first user for uniquely identifying, among the first Web service users, a second user who uses the first Web service that has a relationship in the first Web service with the first user. Receiving means for receiving, from a first server that provides the first Web service, a second calculated recognition tag obtained by performing the calculation on two recognition tags;
A third computed identification tag obtained by computing a third identification tag for uniquely identifying the first user using the second web service among the users of the second web service; In order to uniquely identify a third user who uses the second Web service having a relationship in the second Web service with the first user among the second Web service users Receiving means for receiving, from a second server that provides the second Web service, a fourth calculated recognition tag obtained by performing the calculation on a fourth recognition tag;
The server apparatus characterized by having. Search means for detecting a match between the first computed recognition tag and the third computed recognition tag;
The first computed recognition tag and the third computed recognition tag are used as the first set of the first Web service user set and the second Web service user set. Storage means for storing in association with a fifth identification tag for uniquely identifying a user of
The second calculated identification tag is used to uniquely identify the second user in the union of the first Web service user set and the second Web service user set. Storage means for storing in association with a sixth recognition tag for
The fourth computed identification tag is a first for uniquely identifying the third user in the union of all users of the first Web service and all users of the second Web service. Storage means for storing in association with the recognition card of 7;
The server device according to claim 7, comprising: Receiving the first additional information in the first Web service of the first user in association with the first computed recognition tag, and receiving the second computed identification tag in the second computed recognition tag The second additional information in the first Web service of the user is received in an associated state, and the third calculated identification tag in the third Web service of the first user Receiving in a state in which additional information is associated, and receiving in a state in which the fourth additional information in the second Web service of the third user is associated with the fourth computed recognition tag. The server device according to claim 8. Receiving means for receiving a request including the first computed recognition tag from the first server;
Search means for searching a set of calculated recognition tags received from the first server and the second server using the received first calculated recognition tag as a search key;
The server device according to claim 9, wherein the seventh recognition tag associated with the fourth computed recognition tag obtained as a search result is transmitted to the first server device. Receiving means for receiving a request including the first computed recognition tag from the first server;
Search means for searching a set of calculated recognition tags received from the first server and the second server using the received first calculated recognition tag as a search key;
The server apparatus according to claim 9, wherein the fourth additional information associated with the fourth computed recognition tag obtained as a search result is transmitted to the first server. A first calculation result obtained by performing a first calculation on a first identification tag for uniquely identifying a first user who uses the first Web service among the users of the first Web service; In order to uniquely identify a second user who uses the first Web service having a relationship with the first user in the first Web service among the first Web service users A second calculated recognition tag obtained by performing the first calculation on a second identification tag, and a third using the first Web service having a relationship in the first Web service with the second user. A third calculated identification tag obtained by performing the first calculation on a third identification tag for uniquely identifying the user of the first Web service among the users of the first Web service, and the first A first familiarity coefficient between the user and the second user, the second user and the third user Receiving means for receiving a second familiarity coefficient between the user and the first server that provides the first Web service;
A computing means for performing a second computation on the first intimacy coefficient and the second intimacy coefficient to calculate a third intimacy coefficient;
A server apparatus comprising: a transmission unit configured to transmit the third closeness coefficient to a second server of a second Web service. The first calculation result obtained by performing the first calculation on the first identification tag for uniquely identifying the first user using the first Web service among the users of the first Web service And a second user who uses the first Web service having a relationship with the first user in the first Web service, and uniquely identifies the first Web service user among the first Web service users Receiving means for receiving, from a first server that provides the first Web service, a second computed recognition form obtained by performing the first computation on a second recognition form;
A third identification tag for performing the first calculation on a third identification tag for uniquely identifying the second user who uses the second Web service among the users of the second Web service The calculation result and the third user who uses the second Web service having a relationship in the second Web service with the second user are uniquely identified among the second Web service users. For this purpose, a fourth calculated recognition tag obtained by performing the first calculation on a fourth recognition tag and additional information in the second Web service of the third user are used as the second Web. Receiving means for receiving from a second server providing the service;
A computing means for performing a second computation on the first intimacy coefficient and the second intimacy coefficient to calculate a third intimacy coefficient;
A server apparatus comprising: a transmission unit configured to transmit the additional information and the third closeness coefficient to the first server. The first calculation result obtained by performing the first calculation on the first identification tag for uniquely identifying the first user using the first Web service among the users of the first Web service And a second user who uses the first Web service having a relationship with the first user in the first Web service, and uniquely identifies the first Web service user among the first Web service users , A second calculated recognition tag obtained by performing the first calculation on a second recognition tag, and a first familiarity coefficient between the first user and the second user, Receiving means for receiving from a first server providing a first Web service;
A third identification tag for performing a first identification on a third identification tag for uniquely identifying the second user using the second Web service among the users of the second Web service; The calculation result and the third user who uses the second Web service having a relationship in the second Web service with the second user are uniquely identified among the second Web service users. For this purpose, a fourth calculated recognition tag obtained by performing the first calculation on the fourth recognition tag, additional information in the second Web service of the third user, and the second user Receiving means for receiving a second familiarity coefficient between the second user and the third user from a second server providing the second Web service;
A computing means for performing a second computation on the first intimacy coefficient and the second intimacy coefficient to calculate a third intimacy coefficient;
The server apparatus characterized by having a transmission means which transmits the said additional information to a said 1st server. A first calculation result obtained by calculating a first identification tag for uniquely identifying a first user using the first Web service among the users of the first Web service; A second for uniquely identifying, among the first Web service users, a second user who uses the first Web service that has a relationship with one user in the first Web service A second calculated recognition tag obtained by performing the calculation on the identification tag, and a third user using the first Web service having a relationship with the second user in the first Web service, The first Web service is provided with a third calculated recognition tag obtained by performing the above calculation on a third recognition tag for uniquely specifying among the users of the first Web service. Receiving means for receiving from the server;
A fourth calculation result obtained by calculating a fourth identification tag for uniquely identifying the first user who uses the second Web service among the users of the second Web service; A fifth for uniquely identifying, among the second Web service users, a fourth user who uses the second Web service that has a relationship in relation to the first user and the second Web service. A fifth computed recognition tag obtained by performing the computation on the recognition tag, and a third user who uses the second Web service having a relationship with the fourth user in the second Web service, A second Web service that provides the second Web service with a sixth calculated recognition tag obtained by performing the calculation on a sixth identification tag for uniquely specifying among the users of the second Web service Receiving means for receiving from the server,
The server apparatus characterized by having. The first computed recognition tag and the fourth computed recognition tag are included in the union of the set of users of the first Web service and the set of second Web service users. Storage means for storing in association with a seventh identification tag for uniquely identifying one user;
The third computed recognition tag and the sixth computed recognition tag are included in the union of the set of users of the first Web service and the set of second Web service users. Storage means for storing in association with a first identification tag for uniquely identifying the three users;
The server apparatus according to claim 15, comprising: Search means for searching for the second computed recognition tag received from the first server and the fifth computed recognition tag received from the second server;
A calculation means for calculating the number of calculated recognition votes obtained by the search;
Transmitting means for transmitting the number of calculated recognition votes to the first server;
The server apparatus according to claim 15, comprising: The server device according to claim 5, wherein the calculation has irreversibility. 15. The server device according to claim 12, wherein the first calculation has irreversibility.