KR101373927B1 - Social network server and method for creating social reputation model using the same - Google Patents

Abstract

Disclosed are a social network server and a method for generating a social reliability model using the same. A method of generating a social confidence model using reliability between users constituting the disclosed social network may be performed by using a connection relationship between users who share each of n (n is an integer of 2 or more). Calculating a consensus; And calculating reliability among the users by using consensus on one or more attributes shared among the users among the n attributes, wherein the attributes are extracted from the profiles of the users. Characterized by inclination. According to the present invention, it is possible to reinforce a virtual connection relationship formed between users having the same property in the existing connection relationship between users through the social trust model, and solve the leanness problem in the social trust model.

Description

SOCIAL NETWORK SERVER AND METHOD FOR CREATING SOCIAL REPUTATION MODEL USING THE SAME}

Embodiments of the present invention relate to a social network server and a method for generating a social trust model using the same, and more particularly, to a social network server for generating a social trust model using attribute information of users and a method for generating a social trust model using the same. will be.

A social network is a social structure composed of relationships between individuals or groups formed by one or more interdependent relationships, and can be expressed using a node called a user and a link representing a relationship between them. Here, the relationship between users means a connection relationship between nodes based on structural attributes or user information.

Social networks based on these social networks have made many advances due to the advent of Web 2.0, which emphasizes user participation and collaboration in recent years. Social media refers to online tools and platforms that are used to share multiple opinions, thoughts, experiences, and perspectives.

Content shared through social media can take many forms, including text, images, audio, and video, and include blogs, social networks (Facebook, Twitter, etc.), instant message boards, podcasts, wikis, and user-generated content (UCC). The shared objects have various forms such as text, images, audio, and video.

Under these circumstances, it is important to build a social trust model of a trusted user for searching and recommending social media using a social network. The reliability model refers to a model that shows reliability information on how much the user who is the target of recommendation has the information requested by the recommendation requester.

The existing researches on search and recommendation systems using social networks have used the method of building social trust model by using the connection between users. However, since the number of actual connections between users is limited, the sparsity problem of social networks occurs when a social network is created using the connections between users and a reliability model is built based on the connections.

가 된다. 하지만, 고립된 사용자 및 사용자 집단으로 인해서, 실제로 구성되는 소셜 네트워크는 노드 간 링크의 수가 최대 수 대비 5%로 이하로 존재하게 되는 희박성 문제를 가지게 된다. 그 결과 소셜 네트워크를 기반으로 하는 다양한 추천 및 검색 시스템에서 제공되는 정보에 대한 신뢰성(Reliability) 감소 문제가 발생하게 된다.For example, if we define a set of n finite users as A = {a1, a2, a3, ..., an}, the maximum number of possible links in this group is theoretically

. However, due to isolated users and user populations, the social networks that are actually constructed have a leanness problem where the number of links between nodes is less than 5% of the maximum number. As a result, there is a problem of reducing reliability of information provided by various recommendation and search systems based on social networks.

In order to solve the problems of the prior art as described above, the present invention proposes a novel reliability model generation system and method for extending a social network by using the possibility of information sharing by common attributes among users.

Other objects of the invention will be apparent to those skilled in the art from the following examples.

According to a preferred embodiment of the present invention to achieve the above object, a method for generating a social confidence model using the reliability between users constituting the social network, n (n is an integer of 2 or more) each of the attributes Calculating a consensus on each of the n attributes by using a connection relationship among shared users; And calculating reliability among the users using consensus of one or more attributes shared among the users among the n attributes, wherein the attributes are personal propensity of the users that can be extracted from the profiles of the users. Provided is a method for generating a social reliability model, characterized in that.

The connection relationship includes the joint creation of content among the users, the addition of favorites for content written by other users, the sharing of the same content among the users, tagging on blogs, following on Twitter, and posts. It may include at least one of a comment on, the transmission of the email.

The calculating of the sympathy may include: a bias degree of the connection relationship calculated through the number of connection relationships between users who share any one of the n attributes and the users constituting the social network ( fraction) to calculate the consensus of any one of the above attributes.

The calculating of the sympathy may further include using a clustering correlation coefficient value calculated through a createable connection relationship between users constituting the social network and a createable connection relationship between users who share one of the attributes. The sympathy of any one of the above attributes can be calculated.

The calculating of the reliability may include calculating each of the reliability between the users by using the sum of the consensus on at least one attribute shared by the users.

The method may further include generating a social trust model using the trust between the users, wherein the social trust model of any one of the users may be a trust set of users who have a connection relationship with the user. have.

The attribute may include at least one of age, nationality, gender, occupation, blood type, and address of the users.

According to still another embodiment of the present invention, in a social network server that generates a social reputation model by using reliability between users constituting the social network, among users sharing each of n (n is an integer of 2 or more) attributes A consensus calculation unit for calculating a consensus on each of the n attributes using a connection relationship of? And a reliability calculator configured to calculate reliability between the users using consensus of one or more attributes shared among the users among the n attributes, wherein the attributes are extracted from the profile of the users. There is provided a social network server, characterized by personal disposition.

According to the present invention, it is possible to reinforce a virtual connection relationship formed between users having the same property in the existing connection relationship between users through the social trust model, and solve the leanness problem in the social trust model.

1 is a diagram illustrating a hierarchical structure of personal information of a social network according to an embodiment of the present invention.
2 is a diagram illustrating an attribute according to an embodiment of the present invention.
3 is a diagram illustrating a social confidence model generation system according to an embodiment of the present invention.
4 is a block diagram illustrating a detailed configuration of a social network DB according to an embodiment of the present invention.
5 is a diagram illustrating a detailed configuration of a social network server according to an embodiment of the present invention.
6 is a flowchart illustrating the entire process of the method for generating a social confidence model according to an embodiment of the present invention.

While the invention is susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and will herein be described in detail. It should be understood, however, that the invention is not intended to be limited to the particular embodiments, but includes all modifications, equivalents, and alternatives falling within the spirit and scope of the invention. Like reference numerals are used for like elements in describing each drawing.

Hereinafter, embodiments according to the present invention will be described in detail with reference to the accompanying drawings.

Each user, or group of users, in a social network is defined in scope and character by their attributes.

1 is a diagram illustrating a hierarchical structure of personal information of a social network according to an embodiment of the present invention.

Referring to FIG. 1, an object of each social network, that is, a user has sub-objects indicating their dispositions, and sub-objects have detailed attributes indicating the object's dispositions. .

Therefore, the relationships between users can be expressed in small semantic unit groups such as classes, clusters, etc. according to dependent objects and attributes. Here, the attribute refers to a criterion that reveals a user's personality such as occupation, age, nationality, blood type, gender, address, etc., which can be extracted from the user profile.

2 is a diagram illustrating an attribute according to an embodiment of the present invention.

Referring to FIG. 2, the first user 201 is a woman having a B occupation in 30s. This means that the first user 201 belongs to a temporal attribute of thirties and a professional attribute of B. On the other hand, the second user 202 is a male having a B job in his 30s. Thus, the first user 201 and the second user 202 have a consensus by sharing the same occupation of the same age group. It may also have distinctive characteristics due to different genders.

In other words, users can share their characteristics through attributes that they have in common, and in the present invention, the possibility of information sharing by such common attributes is defined as a trust model.

In order to recognize the collective propensity of users having similar attributes, the present invention analyzes the connection relationship between users who share the same attributes, and calculates the reliability between users. Reliability refers to the possibility of sharing information among users.

More specifically, the present invention calculates the consensus of each of the attributes, and calculates the reliability among the users by using the consensus of the attributes shared among the users.

Here, the sympathy of the attribute refers to the degree of homogeneity or homogeneity of users sharing a specific attribute in the social network. Therefore, high sympathy means that the users who share the property have a high degree of cohesion or association. Therefore, the consensus can be calculated using the degree of connection or sharing between the attributes of the users.

3 is a diagram illustrating a social confidence model generation system according to an embodiment of the present invention.

As shown in FIG. 3, the social confidence model generation system may include a social network server 300 and a social network DB 310, and the social network server 300 includes a plurality of user terminals 330 via a network. It can be connected with.

Here, the network may include both a wired network including the Internet and a dedicated line, and a wireless network such as a broadcasting network and a satellite communication network.

The social network DB 310 stores information on the social network, and stores the reliability between users calculated by the social network server 300 using the attribute information received from the user terminals 320.

4 is a block diagram illustrating a detailed configuration of a social network DB according to an embodiment of the present invention. The function of each component will be described below.

Referring to FIG. 4, the social network DB 210 may include a user storage unit 410, a connection relationship storage unit 420, and a reliability storage unit 430.

The user storage unit 410 stores information such as IDs, passwords, attributes, and the like for the users who form the social network. The stored information may be received from the user terminals 320.

Here, the social network may appear in various forms such as a blog, a facebook, a twitter, a UCC, or a cyworld, and a user may refer to a user subscribed to the service.

The connection relationship storage unit 420 stores information about connection relationships of users who form a social network. A social network is formed by each user as a node, and each user is formed as a connection relationship, and the connection relationship can be represented by a direct connection or an indirect connection.

According to an example of the present invention, the social network server 300 may receive information on connection relations between users through the connection relation storage unit 420.

Here, the connection relationship between users can be described by two relations of unidirectional or both directions. The unidirectional linkage is the case where the information flows in one direction, such as tagging on blog, following on Twitter, unicorn relationship in cyworld, comment on post, sending e-mail.

The two-way connection is a case where the relationship between the users occurs in two directions, such as jointly generating content, adding a link to content written by another user, and sharing the same content. As an example, the co-author relationship of a thesis in a DBLP and the joint link of a thesis refer to a connection relationship in both directions.

It should be apparent to those skilled in the art that the connection relationships in the present invention are not limited to those described above and that there are other connection relationships.

The trust storage unit 430 stores the trust between users calculated using the attributes of the users in the social network server.

Here, reliability refers to the possibility of sharing information among users or the degree of personal trust among users, and may be used as a measure for determining whether to grant access to own information.

According to an embodiment of the present invention, the user storage unit 410, the connection relationship storage unit 420, and the reliability storage unit 430 may be continuously updated when new information is generated. For example, when a new new user is created or a connection relationship between users is added, it may be updated in real time.

5 is a diagram illustrating a detailed configuration of a social network server according to an embodiment of the present invention.

Referring to FIG. 5, the social network server 300 that calculates reliability between users may include a consensus calculation unit 510, a reliability calculation unit 520, and a reliability model generator 530. The function of each component will be described below.

The consensus calculation unit 510 calculates a consensus on each of the n attributes by using a connection relationship between users who share each of n (n is an integer of 2 or more).

In more detail, the consensus calculation unit 510 may calculate the consensus for each attribute by using the associations, that is, the connection relation bias, which users share each attribute. At this time, the consensus calculation unit 510 may refer to the attributes of the users stored in the user storage unit 410 and the connection relationship storage unit 420 and information on the connection relationship of the users.

In more detail, the connection bias value for any one of the n attributes is calculated based on the number of connections among users constituting the social network and connections of users who share any one of the n attributes. Can be.

For example, if there are 100 users constituting a social network, the connection relationship bias for any one property is calculated by using the connection relationship between users who share one property compared to the connection relationship between 100 users. This can be expressed as Equation 1 below.

[ Equation 1 ]

는 p번째 속성의 연결관계 편중도, USER는 사용자, E는 소셜 네트워크를 구성하는 사용자들간의 모든 연결관계 수,

는 사용자가 가지는 p번째 속성을 의미한다.here,

Is the connection bias of the p-th attribute, USER is the user, E is the number of all the connections between users in the social network,

Is the p-th attribute of the user.

According to an embodiment of the present invention, the consensus calculation unit 510 may calculate the consensus for each attribute by further considering the clustering correlation coefficient value of the group of users sharing the attribute.

In more detail, the clustering correlation coefficient value for any one of the n attributes is calculated through all the connectable relations between the users constituting the social network and the createable relations among the users sharing any one attribute. Can be.

For example, assuming that users are randomly distributed, the number of all possible connections to the number e of users who share the same attribute may be expressed as in Equation 2 below.

& Quot; (2 ) & quot ;

는 e명의 사용자들간의 생성 가능한 모든 연결관계의 수를 의미한다. 따라서, 클러스터링 상관 계수 값은 하기의 수학식 3과 같이 표현할 수 있다.here,

Is the number of all possible connections between e users. Therefore, the clustering correlation coefficient value can be expressed as Equation 3 below.

&Quot; (3) "

는 p번째 속성의 클러스터링 상관 계수 값,

는 p번째 속성의 i 번째 세부 속성을 공유하는 사용자들의 수, U는 소셜 네트워크를 구성하는 사용자들의 수를 각각 의미한다.here,

Is the clustering correlation coefficient value of the pth attribute,

Denotes the number of users sharing the i-th detailed attribute of the p-th attribute and U denotes the number of users constituting the social network, respectively.

For example, the p-th attribute is age, and the detailed attributes for age attribute are teenagers, 20s, 30s, 40s, and 50s or more. 23 users, 37 users with 20 detail attributes, 15 users with 30 detail attributes, 17 users with 40 detail attributes, 17 users with more than 50 detail attributes Equation 3, which calculates the sympathy with age in case of 8 users

와 같이 표현될 수 있다.

Can be expressed as

The consensus calculation unit 510 calculates the consensus of each attribute by using the connection relation bias and the clustering correlation coefficient value, which can be expressed as Equation 4 below.

&Quot; (4) "

는 p 번째 속성의 공감도를 의미한다.here,

Denotes the consensus of the p th attribute.

Therefore, if the sympathy of the attribute is large, it means that the connection relationships of users sharing the attribute are closely related to the attribute.

The reliability calculator 520 calculates the reliability between users by using the consensus of one or more attributes among the n attributes.

In more detail, the reliability of the users may be calculated using the sum of the consensus on at least one attribute shared by the users.

For example, if the user A and the user B share the a1, a2, and a7 attributes, the reliability calculator 520 calculates the confidence value between the user A and the user B through the sum of the consensus values of the a1, a2, and a7 attributes. can do.

Therefore, even if the number of attributes shared between users is the same, the reliability between users may be calculated differently according to the magnitude of the consensus of the attributes.

The reliability model generator 530 generates a social confidence model by using the reliability between users calculated by the reliability calculator 520.

Herein, the social confidence model of any one of users refers to a confidence set of users who have a connection with any one of the users.

The present invention can reinforce the virtual connection relationship formed between users having the same property in the existing connection relationship between the users formed through the social trust model generated as described above, and thus can solve the leanness problem in the social trust model. have.

6 is a flowchart illustrating the entire process of the method for generating a social confidence model according to an embodiment of the present invention.

First, in step S600, the consensus calculation unit 510 calculates the consensus of each of the n attributes by using a connection relationship between users who share each of n (n is an integer of 2 or more).

According to an embodiment of the present invention, the consensus calculation unit 510 may calculate the consensus of each attribute by using values of connection relation bias and clustering correlation coefficient of users sharing the respective attributes.

Subsequently, in step S605, the reliability calculator 520 calculates the reliability between the users by using the consensus on one or more attributes among the n attributes.

More specifically, the reliability of the users may be calculated using the sum of the consensus of one or more attributes shared among the users.

Finally, in step S610, the reliability model generator 530 generates a social confidence model using the reliability between users.

So far, embodiments of the method for generating a social trust model according to the present invention have been described, and the configuration of the social network server 300 described with reference to FIG. 5 may be applied to the present embodiment as it is. A detailed description thereof will be omitted.

In addition, embodiments of the present invention may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer readable medium. The computer-readable medium may include program instructions, data files, data structures, and the like, alone or in combination. The program instructions recorded on the medium may be those specially designed and constructed for the present invention or may be available to those skilled in the art of computer software. Examples of computer-readable media include magnetic media such as hard disks, floppy disks, and magnetic tape; optical media such as CD-ROMs and DVDs; magnetic media such as floppy disks; Examples of program instructions, such as magneto-optical and ROM, RAM, flash memory and the like, can be executed by a computer using an interpreter or the like, as well as machine code, Includes a high-level language code. The hardware devices described above may be configured to operate as at least one software module to perform operations of one embodiment of the present invention, and vice versa.

As described above, the present invention has been described by specific embodiments such as specific components and the like. For those skilled in the art, various modifications and variations are possible from these descriptions. Accordingly, the spirit of the present invention should not be construed as being limited to the embodiments described, and all of the equivalents or equivalents of the claims, as well as the following claims, belong to the scope of the present invention .

201: first user 202: second user
300: social network server 310: social network DB
320: user terminal
410: user storage unit 420: connection relationship storage unit
430: reliability storage unit
510: sympathy calculator 520: reliability calculator
530: reliability model generator

Claims (12)

A method of generating a social reliability model using the reliability between users constituting a social network through a social network server including a consensus calculation unit and a reliability calculation unit,
Calculating, by the consensus calculation unit, a consensus degree for each of the n attributes by using a connection relationship between users who share each of n (n is an integer of 2 or more)-any one of the n attributes The sympathy of any one property is calculated by using the fraction of the connection relationship calculated through the connection relationship between users who share the property of and the number of connection relationships between users constituting the social network. -And
Comprising the step of calculating, in the reliability calculator, a reliability representing the possibility of information sharing between the users by using the sum of the consensus of one or more attributes shared among the users among the n attributes,
The attribute means a social confidence model generation method characterized in that the personality of the user extractable from the profile of the user. The method of claim 1,
The connection relationship is the joint creation of content between the users, add bookmarks for content authored by other users, sharing the same content between the users, tagging on blogs, following on Twitter, posts Social confidence model generation method comprising at least one of a comment on, the transmission of the email. delete The method of claim 1,
The step of calculating the consensus,
The sympathy of any one attribute is further obtained by further using a clustering correlation coefficient value calculated through a createable connection relationship between users constituting the social network and a createable connection relationship between users who share the one attribute. Social reliability model generation method characterized in that the calculation. delete The method of claim 1,
The social network server further includes a reliability model generator,
The social confidence model generation method may further include generating a social confidence model by using the reliability between the users in the reliability model generator,
The social confidence model of any one of the user is a social confidence model generation method, characterized in that the set of confidence of the user having a connection with the any one user. The method of claim 1,
The attribute may include at least one of age, nationality, gender, occupation, blood type, and address of the users. In the social network server for generating a social reliability model using the reliability between the users constituting the social network,
A consensus calculation unit that calculates a consensus of each of the n attributes by using a connection relationship among users who share each of n (n is an integer of 2 or more) attributes-share any one of the n attributes Calculating a consensus of the one attribute by using a fraction of a connection relationship calculated through a connection relationship between users and a number of connection relationships between users constituting the social network; And
A reliability calculator configured to calculate a reliability indicating a possibility of sharing information among the users using a sum of consensus on one or more attributes shared among the users among the n attributes,
And the property means personal preferences of the users that can be extracted from the profiles of the users. delete 9. The method of claim 8,
The consensus calculation unit,
The sympathy of any one attribute is further obtained by further using a clustering correlation coefficient value calculated through a createable connection relationship between users constituting the social network and a createable connection relationship between users who share the one attribute. Calculating a social network server. delete 9. The method of claim 8,
Further comprising a reliability model generator for generating a social reliability model using the reliability between the user,
The social trust model of any one of the users is a social network server, characterized in that the set of confidence of the user having a connection with any one user.

Images