KR101243289B1 - Information searching system for social network Service - Google Patents

Abstract

In the social network service information retrieval system according to the present invention, a reference user selection unit for selecting a reference user from among users of a social network service through a network, and collecting reference user information for collecting social network usage information of the reference user selected through the network The reference user score calculator which calculates a reference user score by analyzing social network usage information of the selected reference user, and a peripheral device that identifies neighboring users who have a relationship with the reference user and the social network from the selected reference user's social network usage information. The user analysis unit, the peripheral user score calculator that analyzes the social network usage information of the selected reference user and calculates the peripheral user scores of each of the identified surrounding users, and the social network usage information and the reference user score of the selected reference user. It includes a database unit for storing the individual user information of the selected reference user, and stores the peripheral user score as individual user information of each peripheral user, wherein the user selection unit is selected according to the individual user information of the peripheral users stored in the database unit Selected neighbor users as new reference users and updated.

Description

Information search system for social network service

The present invention relates to a social network service information retrieval system and a social network service information retrieval system for efficiently retrieving social network service information based on social network usage information of users on a social network service.

As social network services develop and the number of users using them increases, the desire of users to retrieve information on social networks also increases.

Information on social networks is usually achieved through collectors, which are automated collection robots. The information on the social network is streaming information that occurs continuously. In order to retrieve the information, the information continuously generated must be collected and stored at all times. However, as the users of the social network and the information continuously occurring increase, it is difficult to collect the information on the social network.

In order to solve the above problems, the present invention provides a social network service information retrieval system capable of efficiently collecting and retrieving information on a social network.

The social network service information retrieval system according to an exemplary embodiment of the present invention may include a reference user selector configured to select a reference user from among users of a social network service through a network, and the social network use information of the reference user selected through the network. A reference user information collection unit to collect, a reference user score calculation unit configured to analyze the selected social network usage information of the reference user, and to calculate a reference user score, from the social network usage information of the selected reference user to the reference user and the social network Peripheral user analysis unit to identify the neighbors with a relationship formed, a neighboring user score calculation unit for analyzing the social network usage information of the selected reference user to calculate the neighboring user score of each of the identified peripheral users and the selected criteria And a database unit for storing the user's social network usage information and the reference user score as individual user information of the selected reference user, and storing the peripheral user score as individual user information of each peripheral user, wherein the user selection unit And selecting and updating the selected neighboring user as a new reference user according to the individual user information of the surrounding users stored in the database unit.

The user selector may select and update a new reference user according to the individual user information stored in the database unit every set time.

The reference user score calculator selects a higher one of the updated reference user score of the updated reference user and the peripheral user score of the updated reference user stored in the database unit as the reference user score of the updated reference user. The database may be stored as individual user information of the updated reference user.

The reference user score calculator may calculate the reference user score by summing the weights of the selected activity cycle scores of the reference user, the selected relationship score of the reference user, and the reference scores of the selected reference user.

The neighboring user score calculator calculates the neighboring user score by summing the predicted score of the neighboring user and the weighted value of each of the selected reference user's reference scores, wherein the predicted score of the neighboring user is the social network of the selected reference user. The reference frequency of the surrounding user in the usage information and the reference user score can be calculated.

The reference score of the selected reference user is calculated from the frequency of mentioning the selected reference user in individual user information other than the individual user information of the selected reference user among the individual user information stored in the database unit, The mention frequency rate of the neighboring user is calculated from the frequency of mentioning the neighboring user in the selected social network usage information of the reference user, and the mentioned frequency of the selected reference user or the mentioned frequency of the neighboring user is selected. It can be obtained from the name of the reference user or the surrounding user, the name of the screen, the ID, the number of appearance of the e-mail address, and the number of times of delivering the written message of the reference user.

Query query analysis unit for analyzing a query input from the outside, a query search unit for searching the search information associated with the query among the individual user information stored in the database using the analysis result of the query and the search information The apparatus may further include an information search unit including a search result output unit output to the outside.

The social network service information retrieval system according to the present invention calculates a user score to first identify a user of high importance, such as a high activity in a social network service or a relationship with other users, to collect information focused on a user of high importance. The order in which the information is collected can be accessed by the user. Therefore, by using the same resources to collect information of high importance or high priority, information on social network services can be searched quickly and accurately.

In addition, since the user may not collect information about a user who has a low activity or a low relationship among a large number of users on the social network service, the efficiency of managing and using the collected information can be improved.

1 is a conceptual diagram illustrating a relationship of users in a social network service (SNS) according to an embodiment of the present invention.
2 is a schematic diagram illustrating a configuration of a social network service information retrieval system according to an embodiment of the present invention.
3 is a schematic diagram showing a configuration of a ranking apparatus according to an embodiment of the present invention.
4 is a schematic diagram illustrating a configuration of an information search unit according to an exemplary embodiment.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. However, the following embodiments of the present invention may be modified into various other forms within the scope of the same invention, and the scope of the present invention is not limited to those described in the following embodiments and accompanying drawings. In the following description, when a component is described as being connected to another component, it may be directly connected to another component, and a third component may be interposed therebetween. In addition, in the drawings, the shape or size of each component is exaggerated for convenience and clarity of description, and parts irrelevant to the description are omitted. Wherein like reference numerals refer to like elements throughout. It is to be understood that the terminology used is for the purpose of describing the present invention only and is not used to limit the scope of the present invention.

1 is a conceptual diagram illustrating a relationship of users in a social network service (SNS) according to an embodiment of the present invention.

Referring to FIG. 1, the social network service 100 includes a plurality of users 10. Each user 10 may form a relationship 20 with some of the other users 10. In the social network service 100, one user 10 may form a relationship 20 with another user 10, such as twitter, facebook, cyworld, blog, and the like. All services on the network may be included. The user 10 may be represented in a manner that can identify the user 10 such as a name, a screen name (alias), an ID (ID, identification number), an e-mail address, and the like. The relationship 20 may be directional or formed bidirectionally with names such as friends / colleagues / followers / neighbors.

One or a plurality of users 10 may be selected as reference users A1 and A2. The user 10 forming a direct relationship 20 with the selected reference users A1 and A2 may be identified as the peripheral users a1 and a2 of the selected reference users A1 and A2. The reference users A1 and A2 may be selected by analyzing the information on the social network service collected in advance. If there is no information on the novel network service collected in advance or there is little information, any user 10 may be selected as the reference user A1 or A2. As will be described later, social network usage information of the selected reference users A1 and A2 may be collected. The social network usage information is a social network service such as posts, links, pictures / pictures, music, videos, etc., which are written or registered by the user 10 in the social network service 100, for example, messages, comments, and posts. Log information such as accessing or visiting 100, reading, writing, and moving may be included. The social network usage information may be streaming information that occurs continuously.

After collecting the social network usage information of the reference users A1 and A2, a new reference user B1 and B2 is assigned to one or a plurality of peripheral users a1 and a2 for each of the selected reference users A1 and A2. Can be selected and updated. Thereafter, the neighboring users b1 and b2 are similarly identified with respect to the updated reference users B1 and B2 and the social network usage information of the updated reference users B1 and B2 may be collected. In this case, although the direct relationship 20 is formed with the updated reference users B1 and B2, the neighboring reference users b1a and b2a who have already been selected as the reference users A1 and A2 and collected the social network usage information are provided. The social network usage information may not be collected.

This process is continuously performed, and by selecting and updating a new reference user every set time, the above-described process is repeated to newly generate the social network usage information for the user 10 in which the social network usage information has already been collected. Can be collected and updated.

2 is a schematic diagram illustrating a configuration of a social network service information retrieval system according to an embodiment of the present invention.

1 and 2 together, the social network service information retrieval system 1000 may be connected to the social network service 100 through the network 10 to exchange information, and also through the network 10 or A query may be directly input from the queryer 200 to output a search result.

The social network service information retrieval system 1000 includes a reference user information collecting unit 1100, a ranking device 1200, and a database unit 1500. In addition, the social network service information retrieval system 1000 may further include an information retrieval unit 1300.

The reference user information collecting unit 1100 may include a collector which is at least one automatic collecting robot. The reference user information collecting unit 1100 may collect information in the social network service 100 through the network 10. In particular, the reference user information collection unit 1100 may collect the social network usage information of the selected reference user A1 or A2 or the updated reference user B1 or B2. The social network usage information collected by the reference user information collection unit 1100 may be stored in the database unit 1500. The social network usage information of each user 10 stored in the database unit 1500 may be referred to as individual user information. The social network usage information of the reference users A1 and A2 may be referred to as individual user information of the individual users A1 and A2. Individual user information may include a word vector containing information about each user 10. The individual user information may include an indexing word vector for indexing and searching the collected social network usage information and dozens of the social network usage information of each user 10.

The database unit 1500 may be one storage device that is physically divided or a division unit that logically divides one or a plurality of storage devices. Alternatively, the database unit 1500 may be part of one storage device that is physically divided or one division unit that is logically divided. The social network usage information collected by the reference user information collection unit 1100 may be stored as individual user information so as to be distinguished for each user 10 when stored in the database unit 1500.

The ranking apparatus 1200 analyzes the social network usage information of the selected reference users A1 and A2 collected by the reference user information collecting unit 1100 to identify the surrounding users a1 and a2 and to determine the identified surrounding users. One or a plurality of (a1, a2, b1, b2) may be selected and updated as the new reference users B1, B2. The new reference users B1 and B2 selected by the ranking apparatus 1200 are transferred to the reference user information collecting unit 1100 so that the reference user information collecting unit 1100 replaces the existing selected reference users A1 and A2. The social network usage information of the updated reference users B1 and B2 may be collected by updating to the reference users B1 and B2.

The information retrieval unit 1300 analyzes the query word input from the interrogator 200 through the network 10 or through the outside of the social network service information retrieval system 1000 to store the individual stored in the database unit 1500. Search for information related to the query word among user information to form search information to output the search information to the queryer 200 through the network 10 or through the outside of the social network service information search system 1000. Can be.

3 is a schematic diagram showing a configuration of a ranking apparatus according to an embodiment of the present invention.

Referring to FIGS. 1 and 3, the ranking apparatus 1200 includes a reference user selector 1210, a reference user score calculator 1220, a neighbor user analyzer 1230, and a neighbor user score calculator 1240. .

The reference user selector 1210 may include one or more users 10 having a high score among a reference user score and a neighboring user score stored in the database unit 1500, which will be described later, among the users 10 of the social network service 100. May be selected as reference users A1 and A2. Alternatively, the peripheral user scores of the peripheral users a1 and a2 of the selected reference users A1 and A2 are calculated to convert one or a plurality of peripheral users a1 and a2 having a high score into the new reference users B1 and B2. Can be selected. The reference user selection unit 1210 may refer to the user having the highest score or the user having the highest score in consideration of the number of users 10 who can collect social network usage information at the same time by the reference user information collecting unit 1100. It can be selected as the users A1, A2, B1, and B2.

If no user has collected or has a small number of novel network usage information, any user may be selected as the reference user A1 or A2. In this case, repeatedly selecting a random user as the reference user (A1, A2) or randomly selecting some of the neighboring users (a1, a2) until the number of users of the novel network usage information is more than a predetermined number. It can be selected as the reference user A1 or A2.

The reference users A1, A2, B1, and B2 refer to users selected at the "current stage" among the users 10 and may change with time. Therefore, hereinafter, when the member numbers A1, A2, B1, and B2 are not used for the "reference user", it may mean "a reference user selected at this stage." Similarly, when the member numbers a1, a2, b1, and b2 are not used for the "peripheral user," it may mean "a peripheral user of the selected reference user at this stage."

The reference user score calculator 1220 analyzes the reference user's social network usage information collected by the reference user information collecting unit 110 and individual user information of other users stored in the database unit 1500 together to revise the reference user score. Score) can be calculated. The reference user score may be represented by Equation 1 below.

Here, RevisitScore is a reference user score, ActivityFrequencyScore is an activity cycle score, RelationsScore is a relationship score, and MentionedScore is a reference score. α is a part for determining the weight of each score, and may have a value between 0 and 1.

That is, the reference user score (RevisitScore) may be calculated by adding the weights of each of the reference user's activity cycle score (ActivityFrequencyScore), the reference user's relation score (RelationsScore), and the reference user's reference score (MentionedScore).

Activity Frequency Score (ActivityFrequencyScore) is a score representing the activity cycle of the reference user, how many content the reference user registered during the unit time, access to the social network service 100 or visit / move within the social network service 100 It can be calculated based on how many times you have done your back.

RelationshipsScore can be calculated based on the number of users that a user is forming in any form. For example, it may be calculated based on the number of followers / following, the number of friends / countries / fans / neighbors, and the like.

Mention Score (MentionedScore) is a score indicating the degree to which the reference user is mentioned from the individual user information of each of the other users 10 other than the individual user information of the reference user already collected and stored in the database unit 1500. The reference score refers to the reference user indicating the reference user's name, screen name (nickname), ID, the number of appearances of the e-mail address, and the number of delivery of the created message written by the reference user in the individual user information of each of the other users 10. The calculated frequency can be calculated.

The neighbor user analyzer 1230 may discover a user 10 having a relationship 20 formed with the reference user and the social network 100 as the neighbor user. The surrounding user analyzer 1230 may analyze the surrounding user from individual user information of the reference user collected by the reference user information collecting unit 1210.

The neighboring user score calculator 1240 may calculate the neighboring user score by summing the specific gravity values of each of the neighboring user's predicted score and the reference score of the reference user. Therefore, the first user score (FirstVisitScore) may be calculated by analyzing social network usage information of the reference user. The peripheral user score may be represented by Equation 2 below.

Here, FirstVisitScore is the surrounding user score, and PredictScore is the prediction score of the surrounding users. MentionedScore is a score obtained by calculating MentionedScore of Equation 1 with respect to surrounding users. The prediction score (PredictScore) may be represented by Equation 3 below. β is a part for determining the weight of each score, it may have a value between 0 and 1.

Here, the prediction score (PredicScore) is a score for estimating the relationship between the reference user and the neighboring user. RevisitScore is a reference user score calculated in Equation 1 above. LocalMentionedRate is the mention frequency, which is the ratio of the number of times all users are mentioned and the number of mentions of neighboring users in the individual user's information of the reference user. (LocalMentionedRate) has a high value. Therefore, the predicted score (PredicScore) of the neighboring user may be calculated from the reference frequency rate (LocalMentionedRate) of the neighboring user and the reference user score (RevisitScore) in the reference user's social network usage information.

The calculated reference user score (RevisitScore) and the reference user's social network usage information may be stored in the database unit 1500 as individual user information of the reference user, and the calculated peripheral user score (FirstVisitScore) may be stored in the database unit 1500 as individual user information of the surrounding user. Can be. The individual user information stored in the database unit 1500 may include all the collected, calculated and analyzed information related to each user 10, and may be further updated by the collected, calculated or analyzed information. .

The user selector 1210 selects one or some of the neighbor users as a new reference user according to the individual user information of the neighbor users, that is, the neighbor user scores of the neighbor users, and updates the reference user in the reference user information collection unit 1100. You can. The user selector 1210 may update the users having the highest score or the highest score as the new reference user.

The reference user score calculator 1220 calculates the updated reference user score (RevisitScore) of the updated reference user, and compares the higher reference user score (FirstVisitScore) of the updated reference user with the higher reference user score of the updated reference user. You can choose with (RevisitScore). The existing individual user information of the updated reference user stored in the database unit 1500 may be updated by storing the selected reference user score (RevisitScore).

4 is a schematic diagram illustrating a configuration of an information search unit according to an exemplary embodiment.

Referring to FIG. 4, the information search unit 1300 may include a query analyzer 1310, a query search unit 1320, and a search result output unit 1330. The query analysis unit 1310 may analyze a query input through the network 10 or externally and convert the analysis result into a searchable word vector.

The query search unit 1320 may search the search information associated with the query by comparing the analysis result of the query with individual user information stored in the database unit 1330. The search information may include all or part of the individual user information. When the analysis result of the query word is expressed as a word vector, the word vector which can be included in the individual user information and the distance thereof are calculated, thereby searching the search information associated with the query word among the individual user information. .

The search result output unit 1330 may output the search information to the interrogator 200 shown in FIG. 1.

1 through 4, the social network service information retrieval system 1000 uses social networks of the selected reference users A1 and A2 in order to retrieve information of the social network service 100 quickly and efficiently. After collecting the information, the neighboring user having a high neighboring user score among the neighboring users a1 and a2 is updated to the new reference users B1 and B2, and the new reference user B1 and B2 is used again. Collect it. By repeating this method, information of the user 10 having a high reference / peripheral user score may be collected first. In addition, a user having the highest score (a reference user score and a neighboring user score) may be newly set as a reference user from individual user information collected at a predetermined time.

The social network service 100 in which a lot of information is quickly generated by many users 10 may preferentially collect information of users who are likely to be searched by the queryer 200.

In addition, since the information of the user who has little activity or has little relationship with other users can be excluded from the collection, the database unit 1500 for storing and managing the collected information can be efficiently used, and the database unit 1500 The stored information may be quickly and efficiently retrieved by the information retrieval unit 1300.

Claims (7)

A reference user selector configured to select a reference user among users of a social network service through a network;
A reference user information collecting unit which collects social network usage information of the reference user selected through the network;
A reference user score calculator configured to calculate a reference user score by analyzing the selected social network usage information of the reference user;
A peripheral user analyzer configured to identify neighboring users having a relationship with the reference user and the social network from the selected social network usage information of the reference user;
A neighbor user score calculator configured to analyze the social network usage information of the selected reference user and calculate the neighbor user score of each of the identified neighbor users; And
And a database unit for storing the selected social network usage information of the reference user and the reference user score as individual user information of the selected reference user, and storing the surrounding user score as individual user information of each neighboring user. But
The reference user selection unit selects and updates a neighboring user selected according to individual user information of the neighboring users stored in the database as a new reference user,
The reference user score calculator is configured to calculate the reference user score by summing the weights of the selected activity cycle scores of the reference user, the selected relationship score of the reference user, and the reference scores of the selected reference user. Social network service information retrieval system. The method according to claim 1,
The reference user selection unit selects and updates a new reference user according to the individual user information stored in the database unit at predetermined time intervals. The method of claim 2,
The reference user score calculator selects a higher one of the updated reference user score of the updated reference user and the peripheral user score of the updated reference user stored in the database unit as the reference user score of the updated reference user. And storing the updated individual user information of the updated reference user in the database unit. delete The method according to claim 1,
The neighboring user score calculator calculates the neighboring user score by adding the weighted values of each of the predicted scores of the neighboring users and the reference scores of the selected reference user,
And the predicted score of the neighboring user is calculated from the reference frequency rate of the neighboring user in the selected social network usage information of the reference user and the reference user score. 6. The method of claim 5,
The reference score of the selected reference user is calculated from the frequency of mentioning the selected reference user in individual user information other than the individual user information of the selected reference user among the individual user information stored in the database unit,
The mention frequency rate of the neighboring user is calculated from the frequency mentioned by the neighboring user in the selected social network usage information of the reference user,
The mentioned frequency of the selected reference user or the stated frequency of the surrounding user may include the name, screen name, ID, e-mail address of the selected reference user or the surrounding user, and the number of times the reference message is delivered. Social network service information retrieval system, characterized in that obtained from. The method according to claim 1,
A query analysis unit which analyzes a query input from the outside;
A query search unit for searching for search information associated with the query among the individual user information stored in the database unit by using the analysis result of the query; And
And a search result output unit for outputting the search information to the outside.

Images