WO2010102527A1 - 一种sns网络中成员关系圈的提取方法和装置 - Google Patents

一种sns网络中成员关系圈的提取方法和装置 Download PDF

Info

Publication number
WO2010102527A1
WO2010102527A1 PCT/CN2010/070309 CN2010070309W WO2010102527A1 WO 2010102527 A1 WO2010102527 A1 WO 2010102527A1 CN 2010070309 W CN2010070309 W CN 2010070309W WO 2010102527 A1 WO2010102527 A1 WO 2010102527A1
Authority
WO
WIPO (PCT)
Prior art keywords
relationship
information
feature
population
database
Prior art date
Application number
PCT/CN2010/070309
Other languages
English (en)
French (fr)
Inventor
殷宇
蔡耿平
胡海斌
Original Assignee
腾讯科技(深圳)有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority to BRPI1009430A priority Critical patent/BRPI1009430A2/pt
Application filed by 腾讯科技(深圳)有限公司 filed Critical 腾讯科技(深圳)有限公司
Priority to MX2011009462A priority patent/MX2011009462A/es
Priority to RU2011140609/08A priority patent/RU2011140609A/ru
Priority to CA2753096A priority patent/CA2753096A1/en
Publication of WO2010102527A1 publication Critical patent/WO2010102527A1/zh
Priority to US13/230,269 priority patent/US9031972B2/en

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L51/00User-to-user messaging in packet-switching networks, transmitted according to store-and-forward or real-time protocols, e.g. e-mail
    • H04L51/52User-to-user messaging in packet-switching networks, transmitted according to store-and-forward or real-time protocols, e.g. e-mail for supporting social networking services
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00Administration; Management
    • G06Q10/10Office automation; Time management

Definitions

  • the present invention relates to the field of computer network technologies, and in particular, to a method and apparatus for extracting membership relationships in an SNS network. Background technique
  • the development of network instant messaging tools has been accepted by most netizens and has become an indispensable software tool for users.
  • Network instant messaging tools are widely used both in casual entertainment and in the work of users.
  • Instant messaging software offers more and more features and features.
  • the social network service (SNS) formed by online netizens is no longer just a relationship between a single user and a single user, but a one-to-many and many-to-many relationship.
  • the social network includes online users and their relationship networks, which has great value, can accurately search and effectively spread information, and meet the different needs of users and enterprises.
  • the SNS network contains a large number of users and massive relationship data. Therefore, a basic problem to be solved is how to find valuable and interesting information from the massive data of the SNS network. Obviously, the massive users and massive data of the SNS network are not all concerned by individual users or enterprise users. Individual users or enterprise users are interested in the relationship of specific targets.
  • the technical problem to be solved by the present invention is that the social network members cannot be represented against the prior art.
  • the relationship between the relationship and the defects formed by it provides a method and device for extracting the membership relationship in the SNS network.
  • a method for extracting membership relationships in a social network service SNS network including:
  • the step (a) comprises:
  • the attention population database stores feature information of the attention population and the attention population corresponds to a unique identity of the SNS network.
  • the step (b) further includes:
  • the contact information includes: a unique identifier, a feature information, a relationship type, and a relationship weight of the contact corresponding to the SNS network.
  • the step (b2) further includes:
  • step (b22) determining whether the current relationship ⁇ has reached the predetermined value of the scale, and if so, completing the relationship ⁇ extraction step, otherwise performing step (b23);
  • step (b24) extracting feature information of the current relationship ⁇ member; and determining whether the current level is 0, such as If yes, perform step (b26), otherwise perform step (b25);
  • step (b25) determining whether the feature information of the current relationship ⁇ member satisfies the feature filtering condition, and if yes, performing step (b26), otherwise performing step (b22);
  • step (b26) judge whether the current level + 1 is less than the predetermined level, if yes, execute step (b27), otherwise perform step (b22);
  • step (b27) traversing each contact of the current relationship ⁇ member, adding the contact as a relationship ⁇ member to the end of the unprocessed queue when the contact is neither in the unprocessed queue nor the processed queue, and The level of the contact is set to the current level + 1 , and then step (b28) is performed; otherwise, step b (28) is directly executed;
  • step b28 Store the feature information, contact information, and hierarchy of the current relationship member into the relationship database, and then perform step b (22).
  • the method further comprises:
  • the display content includes: feature information of a relationship member, a relationship type, a weight information, and a relationship path.
  • the method further comprises:
  • step (d) further comprises:
  • an apparatus for extracting member relationships in the social network service SNS network including:
  • a target crowd selection module configured to select a target population in the SNS network
  • a relationship extraction module configured to analyze a relationship chain of the target population, and extract a relationship relationship between the target population from a relationship chain of the target population according to a feature filtering condition.
  • the relationship extraction module includes:
  • a relationship obtaining unit configured to acquire contact information of the target group from a relationship chain of the target group;
  • the relationship ⁇ building unit is configured to select a relationship ⁇ member of the target person according to the feature filtering condition and contact information and determine a level at which the relationship ⁇ member is in the relationship ⁇ .
  • the device further comprises:
  • Focusing on a population database configured to store feature information of the focused population and a unique identity of the focused population corresponding to the SNS network;
  • the target crowd obtaining module selects a target group from the database of interested people according to the query condition.
  • the device further comprises:
  • SNS network member data database used to store data of SNS network members
  • Feature rule database for storing feature keywords obtained from the SNS network member profile database
  • a feature extraction module configured to extract network member features from the SNS network member profile database according to the feature keyword;
  • Focusing on the crowd database construction module for selecting a population of interest from the SNS network member data database according to the set feature to construct a database of the attention population;
  • the relationship database is used to store the relationship information of the members of the relationship, the contact information, and the level at which the relationship member is located.
  • the device further comprises:
  • the relationship ⁇ display module is used to display the relationship ⁇ members by level.
  • the device further comprises:
  • the influence calculation module is configured to: select relationship information and/or feature information of each relationship ⁇ member in the relationship ,, and calculate, according to the relationship information and/or feature information, the each relationship ⁇ member in the relationship ⁇ The influence.
  • the invention analyzes the relationship chain of the target population, and extracts according to specific feature filtering conditions.
  • the relationship of the target group is a member and constructs a relationship, thereby providing various network members in the SNS network that meet the specified characteristics and the relationship between them.
  • the present invention can also calculate the influence of the members of the relationship ⁇ in the relationship ⁇ according to the relationship information and/or the feature information of each relationship ⁇ member in the relationship ,, and help to find the most influential relationship ⁇ The crowd makes information transmission and retrieval more targeted.
  • FIG. 1 is a flow chart of a first embodiment of a method for extracting membership relationships in an SNS network according to the present invention
  • FIG. 2 is a flow chart showing a second embodiment of a method for extracting membership relationships in an SNS network according to the present invention
  • 3 is a flow chart of determining the relationship between the target members of the member relationship extraction method in the SNS network of the present invention and its hierarchy;
  • FIG. 4 is a flow chart of a third embodiment of a method for extracting membership relationships in an SNS network according to the present invention.
  • FIG. 6 is a schematic structural diagram of a first embodiment of an apparatus for extracting membership relationships in an SNS network according to the present invention
  • FIG. 7 is a schematic structural diagram of a second embodiment of an apparatus for extracting membership relationships in an SNS network according to the present invention.
  • Figure 8 is a schematic illustration of the relationship of a particular population extracted using an embodiment of the method or apparatus of the present invention.
  • Fig. 9 is a schematic diagram showing the extraction process of the relationship ⁇ of the target population of the present invention. detailed description
  • the relationship chain of the target group can be formed by analyzing the relationship chain of the target group according to the specific feature filtering condition.
  • Network users or distributors can use the relationship of target groups to find valuable relationship chain information, realize accurate search and transmission of information, and facilitate business promotion and cooperation.
  • FIG. 1 is a flow chart of a first embodiment of a method for extracting membership relationships in an SNS network according to the present invention. The specific process is as follows:
  • the target population is first selected in the SNS network.
  • the selection of the target population can be selected from the SNS network members according to specific query conditions, for example, the query condition can be a star of sports or entertainment. It can also be a certain or some designated network member, for example, Beckham, Ronaldo and Ronald Dino can be directly designated as the target group.
  • the relationship chain of the target population is analyzed, and the relationship ⁇ of the target population is extracted from the relationship chain of the target population according to the feature filtering condition.
  • the relationship chain may include all contact information of the target group, contact information between the contact and other SNS network members, and other related information. For example, if the target person A knows B and B knows C, then A and B have direct contact, and A has indirect connection with C. A, B, and C and their related information constitute the most relevant relationship chain of the target person.
  • FIG. 2 is a flow chart of a second embodiment of a method for extracting membership relationships in an SNS network according to the present invention. The specific process is as follows:
  • the attention population database may be established on the server side of the SNS network, and the target population is selected from the population database of interest according to the query condition.
  • the attention population database can be collected from the real world or the media, and can be obtained manually or by the portal's news statistics device or by the search engine's search ranking.
  • the establishment of the attention population database can be implemented in the following manner. First, build an SNS network member database for storing SNS network member data. Then, feature keywords are extracted from the SNS network member database in the self-developed or third-party designed feature statistics rule database. The network member feature is then extracted from the SNS network member profile database according to the feature keywords described above. Finally, according to the set characteristics, data from the SNS network member according to the set characteristics The focus group is selected in the library to build a database of people of interest. In the attention population database, the feature information of the attention population and the unique identity of the attention population corresponding to the SNS network may be stored. The characteristic information of each follower can be expressed as (type, value).
  • user A's favorite movie is a beautiful soul, described as: (favorite movie, beautiful mind).
  • the unique identity corresponding to the SNS network may be a login account of an instant message (IM).
  • Table 1 shows the information that the attention person stores in the attention population database.
  • the selection of the target population can be based on the population database. For example, you can set the query condition to be a world-class star, and select the people who are interested in the condition database, such as Beckham and Ronaldo.
  • relatively complex query conditions can be directly set, and the target population is directly selected in the SNS network member database.
  • the target population can be directly specified.
  • a predetermined value of the relationship ⁇ of the target group, a predetermined value of the hierarchy, and a feature filtering condition of the relationship ⁇ member may be set.
  • the total number of people who can set their relationship is 900
  • the level is 3, and the number of people on each floor is 300 (the number of people on each floor can be the same) Can be different).
  • Feature filtering information can be (professional, player), (other, advertised). Therefore, in this embodiment, it is necessary to find out the relationship of the target group, the person who has taken the advertisement and the occupation is the player.
  • contact information of the target group is acquired from a relationship chain of the target group.
  • the contact information may be a unique identity, feature information, relationship type, and relationship weight of the contact corresponding to the SNS network.
  • the contact information may include a contact list that the target group directly or indirectly contacts using the SNS network. For example, buddy list, blog access user, and so on.
  • the relationship between the target group and the contact can be expressed as (ID, type, value).
  • the ID indicates the unique identity of the contact in the SNS network
  • type indicates the type of relationship, for example, defined as: friend, understanding, stranger.
  • Value is defined as the weight of the relationship, which is the importance of the relationship. The greater the weight, the better the relationship, the more closely the contact, the more frequent.
  • Those skilled in the art can use any known method or material to determine the contact information of the target population.
  • step S204 the relationship ⁇ member of the target person is selected according to the feature filtering condition and the contact information, and the level at which the relationship ⁇ member is located in the relationship ⁇ is determined.
  • step S205 the relationship ⁇ member is displayed in a hierarchy according to the relationship ⁇ database.
  • the display content includes: feature information of a relationship member, a relationship type, weight information, and a relationship path.
  • Fig. 9 shows an extraction process of extracting the relationship ⁇ of the target population of the present invention in accordance with the method shown in Fig. 2.
  • Figure 8 shows a multi-layer relationship.
  • the different styles of the lines in the figure indicate the type of relationship, and further the online values represent the actual weight of the relationship. The greater the weight, the closer the relationship.
  • Each ring represents a level of relationship. There are direct or indirect relationships between each of the two members of the relationship, and the edges of the relationship are connected.
  • Figure 3 shows an exemplary embodiment of step 204 of Figure 2.
  • the current level of the relationship ⁇ is set to 0, and the target group is added as a relationship ⁇ member to the unprocessed queue, and the processed queue is emptied.
  • step S302 it is judged whether the size of the current relationship ⁇ reaches a predetermined value, and if so, the relationship ⁇ extraction step is completed, and the flow ends; otherwise, step S303 is performed.
  • the judgment of the relationship size may include the judgment of the number of members of each layer relationship and the judgment of the total relationship number.
  • step S303 it is determined whether the unprocessed queue is empty. If yes, the relationship extraction step is completed, and the flow ends. Otherwise, step S304 is performed.
  • step S304 the first relationship member is removed from the unprocessed queue as the current relationship member and added to the processed queue, and the current hierarchy is set to the level of the current relationship member.
  • step S305 extracting the feature information of the current relationship ⁇ member, and determining whether the current level is 0, if yes, executing step S307, otherwise performing step S306;
  • step S306 it is determined whether the feature information of the current relationship ⁇ member satisfies the feature filtering condition, and if yes, step S307 is performed, otherwise step S302 is performed;
  • step S307 it is determined whether the current level + 1 is less than the predetermined level, if yes, step S308 is performed, otherwise step S302 is performed;
  • step S308 traversing each contact of the current relationship ⁇ member, when the contact is neither in the unprocessed queue nor the processed queue, step S309 is performed, otherwise step S310 is performed;
  • step S309 the contact is added as a member of the relationship to the end of the unprocessed queue, and the level of the contact is set to the current level + 1, and then step 310 is performed;
  • step S310 the feature information, the contact information and the hierarchy of the current relationship member are stored in the relationship database, and then step S302 is performed.
  • step S302 the relationship information, contact information, and hierarchical storage of the relationship member can be seen in Table 2.
  • FIG. 4 is a flow chart of a third embodiment of a method for extracting membership relationships in an SNS network according to the present invention. The specific process is as follows:
  • the target population is first selected in the SNS network.
  • the selection of the target group can be selected from the SNS network members according to specific query conditions, for example, the query condition can be a star of sports or entertainment. It can also be a certain or some designated network member, for example, you can directly specify Beckham, Ronaldo and Ronald Dino as the target group.
  • step S402 the relationship chain of the target group is analyzed, and the relationship ⁇ of the target group is extracted from the relationship chain of the target group according to the feature filtering condition.
  • step S403 relationship information and/or feature information of each relationship ⁇ member in the relationship ⁇ is selected. And calculating the influence of the respective relationship members in the relationship according to the relationship information and/or the feature information.
  • the relationship information and/or the feature information may be directly obtained from the relationship chain of the target group; or may be obtained from the database of the attention population.
  • the establishment of the population database can be referred to the description above.
  • the characteristic information of the relationship member can be expressed as (type, value). For example, user A's favorite movie is a beautiful mind, described as: (favorite movie, beautiful mind).
  • the method of the present invention according to calculating the influence of the respective relationship members in the relationship, it is helpful to find the most influential people in the relationship, and the information transmission and retrieval are more targeted. .
  • Figure 5 is a typical embodiment of the influence calculation step of Figure 4. As shown in Figure 5,
  • step S501 a relationship ⁇ is selected, and the feature filtering condition of the relationship ⁇ is extracted.
  • step S502 traversing each member in the relationship, selecting feature information of each relationship member, and filtering the matching information according to the relationship between the feature information of the relationship member and the feature of the relationship The feature score of the relationship ⁇ member is calculated.
  • the feature score for each relationship member belonging to a relationship relationship can be calculated as follows:
  • the user can play the information of the network game, including the time and level, to convert the corresponding game points, thereby taking the feature score.
  • each member in the relationship is traversed, the relationship information of the relationship member is selected, and the relationship score of the relationship member is calculated according to the relationship information of the relationship member.
  • the relationship information includes a relationship type, a weight information, and a relationship path between the members of the relationship.
  • the feature scoring function is designed as follows:
  • R (relationship ⁇ member ID) ⁇ type of relationship with each other relationship , member, if it is a friend plus 10 points, know plus 5 points, strange plus 1 point ⁇
  • the feature scoring function is designed as follows:
  • R (relationship ⁇ member ID) ⁇ relationship type with each other relationship * member * weight ⁇ .
  • the weight indicates the importance of the relationship. The greater the weight, the better the relationship, the more closely the contact, the more frequent. For example, when the relationship type is a friend, it counts 10 points, when the relationship type is 5 hours of recognition time, when the relationship type 1 point for a strange time. When the number of contacts in more than 5 times in a month is 3 points, the number of contacts is 3-5 times, 2 points, and less than 3 times, 1 point.
  • step S504 the influence of the relationship ⁇ member is calculated according to the weighted result of the feature score and the relationship score.
  • the influence scoring function is designed as follows:
  • f is the influence weight, the default is 0.5, which can be adjusted according to actual needs.
  • the influence of the relationship ⁇ member can be directly calculated based on the feature information, which will hereinafter be referred to as the feature influence force.
  • the process is as follows:
  • the influence of the relationship member can be directly calculated based on the relationship information, which will be referred to as the relationship influence. The process is as follows:
  • the relationship information includes a relationship type, a weight information, and a relationship path between the members of the relationship.
  • FIG. 6 is a schematic structural diagram of a first embodiment of an apparatus for extracting membership relationships in an SNS network according to the present invention. This embodiment is one of the most compact embodiments of the present invention.
  • the device of the present invention includes: a target crowd selection module 401 and a relationship extraction module 402.
  • the target crowd selection module 401 can be used to select a target population.
  • the selection of the target population may be selected from the SNS network members according to specific query conditions, or the SNS network member may be directly designated by the client.
  • the query condition can be a star of sports or entertainment, or directly specify Beckham, Ronaldo and Ronald Dinho as the target group.
  • the relationship extraction module 402 is configured to analyze a relationship chain of the target population, and extract a relationship ⁇ of the target population from a relationship chain of the target population according to a feature filtering condition.
  • the relationship chain may include all contact information of the target population, contact information between the contact and other SNS network members, and other related information. For example, if the target person A knows B and B knows C, then A and B have direct contact, and A has indirect connection with C. A, B, and C and their related information constitute the most relevant relationship chain of the target person.
  • the member device ⁇ extraction device in the SNS network of the present invention further includes an influence calculation module (not shown).
  • the influence calculation module may be configured to calculate the influence of the respective relationship members in the relationship ⁇ according to relationship information and/or feature information of each relationship ⁇ member in the relationship ⁇ .
  • the relationship information and/or the feature information may be directly obtained from a relationship chain of the target group; or may be obtained from a database of the attention population.
  • the establishment of the population database can be referred to the previous description.
  • FIG. 7 is a block diagram showing the structure of a second embodiment of the apparatus for extracting membership relationships in an SNS network of the present invention.
  • the SNS network member profile database 501 is used to store data for members of the SNS network. This information can be filled out by the network members when they register, or it can be collected by the search engine from the information published by the online media.
  • the information may include: the unique identity of the network member corresponding to the SNS network, the characteristic information, the relationship chain, and the like, any information that has appeared on the network or in real life or has been recorded.
  • the feature rule database 502 is connected to the SNS network member profile database 501 for storing feature keywords acquired from the SNS network member profile database 501.
  • the extraction of the feature keywords can be performed using any of the methods and rules disclosed in the art, and those skilled in the art are familiar with and can skillfully apply the methods and rules.
  • the feature extraction module 503 passes the feature rule database 502 and the The SNS network member profile database 501 is connected to select the feature keywords from the feature rule database 502, and extract network member features from the SNS network member profile database 501 according to the feature keywords.
  • the attention population database construction module 514 can be connected to the SNS network member material database 501, the feature extraction module 503, and the feature rule database 502 to select a population of interest to enter the population of interest database 504 from the SNS network member profile database 501 according to the set characteristics. .
  • the attention population database 504 can also be constructed in other ways.
  • the target crowd selection module 505 can connect to the crowd database 504 and select a target population from it. In other embodiments of the present invention, the target crowd selection module 505 can also directly receive external input or obtain the target population in other manners.
  • the relationship extraction module 506 includes a relationship acquisition unit 5061 and a relationship construction unit 5062.
  • the relationship obtaining unit 5061 can be connected to the target crowd selection module 505 and the SNS network member profile database 501 at the same time, and obtain the target crowd from the target selection module 505, and then according to the target population.
  • the unique identity of the SNS network obtains the relationship chain of the target population from the SNS network member profile database 501 to obtain contact information.
  • the contact information may be a unique identity, feature information, relationship type, and relationship weight of the contact corresponding to the SNS network.
  • the contact information may include a list of contacts directly or indirectly contacted by the target group using the SNS network. For example, IM buddy list, blog access user, and so on.
  • the relationship between the target population and the contact can be expressed as (ID, type, value).
  • the ID indicates the unique identity of the contact in the SNS network
  • type indicates the type of relationship, for example, defined as: friend, understanding, stranger.
  • Value is defined as the weight of the relationship, that is, the importance of the relationship. The greater the weight, the better the relationship, the closer the connection, the more frequent.
  • the target crowd selection module 505 acquires a relationship chain while selecting a target group.
  • the relationship obtaining unit 5061 can obtain the contact information directly from the target crowd selection module 505.
  • the relationship obtaining unit 5061 may also acquire the relationship chain of the target population from other modules (such as the attention population database 504) by using other techniques known in the art to obtain contact information.
  • the relationship ⁇ building unit 5062 can select the relationship ⁇ member of the target person according to the feature filtering condition and the contact information and determine the level at which the relationship ⁇ member is in the relationship ⁇ . Relationship member See Figure 3 for the selection and hierarchy settings. Those skilled in the art can also use other decision processes to accomplish this step in accordance with the teachings of the present invention.
  • the relationship ⁇ display module 508 can be directly coupled to the relationship ⁇ construction unit 5062 to display relationship ⁇ members (not shown) in a hierarchy.
  • the relationship ⁇ constructed by the relationship ⁇ building unit 5062 can be stored in the relation ⁇ database 507.
  • the relationship ⁇ display module 508 can be connected to the relationship ⁇ database 507 (see Figure 7), and the relationship ⁇ members are displayed hierarchically.
  • Each relationship data stored in the relationship database 507 is in the form of a table, and the table may include the ID, characteristic information, contact information, and hierarchical information of each member.
  • the storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Strategic Management (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Human Resources & Organizations (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • General Business, Economics & Management (AREA)
  • Data Mining & Analysis (AREA)
  • Tourism & Hospitality (AREA)
  • Physics & Mathematics (AREA)
  • Signal Processing (AREA)
  • Computing Systems (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

本发明涉及一种 SNS网络中成员关系圈的提取方法和装置。所述方法包括以下步骤:(a)在 SNS网络中选择目标人群;(b) 分析所述目标人群的关系链,并根据特征过滤条件从所述目标人群的关系链中提取所述目标人群的关系圈。所述装置包括目标人群选取模块,用于在 SNS网络中选择目标人群;关系圈提取模块,用于分析所述目标人群的关系链,并根据特征过滤条件从所述目标人群的关系链中提取所述目标人群的关系圈。实施本发明的方法和装置,可提供 SNS网络中满足指定特征的各个人以及他们之间的关系;利用目标人群的关系圈,能够找到有价值的关系链信息,实现信息的精确搜索和传递,方便商业活动推广和合作。

Description

一种 SNS网络中成员关系圏的提取方法和装置 本申请要求于 2009年 3月 12日提交中国专利局、申请号为 200910105978.3、 发明名称为"一种 SNS网络中成员关系圏的提取方法和装置"的中国专利申请的 优先权, 其全部内容通过引用结合在本申请中。 技术领域
本发明涉及计算机网络技术领域, 更具体地说, 涉及一种 SNS网络中成员 关系圏的提取方法和装置。 背景技术
网络即时通讯工具发展到今天, 已经被大多数的网民所接受并且已经成为 用户必不可少的软件工具。 网络即时通讯工具不管是在平时的休闲娱乐中, 还 是在用户的工作中, 都得到广泛的使用。 即时通讯软件提供的功能越来越多, 功能也日趋完善。 与此同时, 在线网民形成的社会性网络服务(Social Networks service, SNS ) 中不再仅仅是单个用户与单个用户的关系, 而是单对多以及多对 多的关系。 社会化网络包含了在线用户及其关系网络, 具有非常巨大的价值, 可以实现信息的精确搜索及有效传播, 满足用户和企业的不同需求。
SNS 网络中包含海量用户及海量关系数据, 因而需要解决的一个基本问题 是如何从 SNS网络的海量数据中找到有价值的、 感兴趣的信息。 显然, SNS网 络的海量用户和海量数据并不是全部为个人用户或企业用户所关注, 个人用户 或企业用户感兴趣的是特定目标的关系圏。
现有的大多数 SNS网站支持通过关键词对网络成员进行搜索, 搜索结果可 以展示出网络中满足指定特征的每个人, 但无法表现这些人之间的关系及其所 形成的关系圏, 并且不能针对特定目标的关系圏进行提取和分析。 因此不能发 现更有价值的关系信息。 发明内容
本发明要解决的技术问题在于, 针对现有技术的无法表现社会化网络成员 的关系及其所形成的关系圏缺陷, 提供一种 SNS网络中成员关系圏的提取方法 和装置。
为了实现发明目的, 提供了所述一种社会性网络服务 SNS网络中成员关系 圏的提取方法, 包括:
(a)在所述 SNS网络中选择目标人群;
(b) 分析所述目标人群的关系链,并根据特征过滤条件从所述目标人群的关 系链中提取所述目标人群的关系圏。
优选地, 所述步骤 (a)包括:
(al)建立关注人群数据库, 并根据查询条件从所述关注人群数据库中选取目 标人群;
(a2)设定所述目标人群的关系圏的规模预定值、 层次预定值和关系圏成员 的特征过滤条件。
优选地, 所述关注人群数据库中存储有所述关注人群的特征信息和所述关 注人群对应于 SNS网络的唯一身份标识。
优选地, 所述步骤 (b)中进一步包括:
(bl)从所述目标人群的关系链获取所述目标人群的联系人信息;
(b2)根据所述特征过滤条件和联系人信息选取所述目标人群的关系圏成员 并判定所述关系圏成员在所述关系圏中所处的层次。
优选地, 所述联系人信息包括: 联系人对应于 SNS网络的唯一身份标识、 特征信息、 关系类型和关系权重。
优选地, 所述步骤 (b2)中进一步包括:
(b21) 将关系圏的当前层次设定为 0, 并将所述目标人群作为关系圏成员添 加到未处理队列中, 同时清空已处理队列;
(b22) 判断当前关系圏的规模是否达到规模预定值,如果是则完成关系圏提 取步骤, 否则执行步骤 (b23);
(b23) 判断未处理队列是否为空, 如果是则完成关系圏提取步骤, 否则从未 处理队列移走第一个关系圏成员作为当前关系圏成员并将其添加到已处理队 列, 并将当前层次设置为所述当前关系圏成员的层次;
(b24)提取所述当前关系圏成员的特征信息; 并判定当前层次是否为 0, 如 果是则执行步骤 (b26) , 否则执行步骤 (b25);
(b25) 判定所述当前关系圏成员的特征信息是否满足特征过滤条件, 如果 是, 则执行步骤 (b26), 否则执行步骤 (b22) ;
(b26) 判断当前层次 + 1是否小于预定层次, 如果是则执行步骤 (b27), 否则 执行步骤 (b22);
(b27)遍历当前关系圏成员的每一个联系人, 当所述联系人既不在未处理队 列也不在已处理队列时将所述联系人作为关系圏成员添加到未处理队列的尾 部, 并将所述联系人的层次设置为当前层次 + 1 , 再执行步骤 (b28); 否则直接执 行步骤 b(28);
(b28) 将当前关系圏成员的特征信息、联系人信息和层次存储到关系圏数据 库中, 再执行步骤 b(22)。
优选地, 所述方法进一步包括:
(c)根据关系圏数据库的信息按照层次显示所述关系圏成员。
优选地, 所述显示内容包括: 关系圏成员的特征信息、 关系类型、 权重信 息和关系路径。
优选地, 所述方法进一步包括:
(d) 选取所述关系圏中各个关系圏成员的关系信息和 /或特征信息, 并根据 所述关系信息和 /或特征信息计算所述各个关系圏成员在所述关系圏中的影响 力。
优选地, 所述步骤 (d)进一步包括:
(dl) 选取关系圏成员的特征信息, 并根据所述关系圏成员的特征信息与所 述关系圏的特征过滤信息之间的匹配程度计算所述关系圏成员的特征评分;
(d2) 选取关系圏成员的关系信息, 并根据所述关系圏成员的关系信息计算 所述关系圏成员的关系评分;
(d3)根据所述特征评分和关系评分的加权结果计算所述关系圏成员的影响 力。
为了更好地实现发明目的, 提供了所述一种社会性网络服务 SNS网络中成 员关系圏的提取装置, 包括:
目标人群选取模块, 用于在所述 SNS网络中选择目标人群; 关系圏提取模块, 用于分析所述目标人群的关系链, 并根据特征过滤条件 从所述目标人群的关系链中提取所述目标人群的关系圏。
优选地, 所述关系圏提取模块包括:
关系获取单元: 用于从所述目标人群的关系链获取所述目标人群的联系人 信息;
关系圏构建单元: 用于根据所述特征过滤条件和联系人信息选取所述目标 人员的关系圏成员并判定所述关系圏成员在所述关系圏中所处的层次。
优选地, 所述装置进一步包括:
关注人群数据库, 用于存储所述关注人群的特征信息和所述关注人群对应 于 SNS网络的唯一身份标识; 并且
所述目标人群获取模块根据查询条件从所述关注人群数据库中选取目标人 群。
优选地, 所述装置进一步包括:
SNS网络成员资料数据库: 用于存放 SNS网络成员的资料;
特征规则数据库: 用于存储从所述 SNS网络成员资料数据库获取的特征关 键词;
特征提取模块: 用于根据所述特征关键词从所述 SNS网络成员资料数据库 提取网络成员特征;
关注人群数据库构建模块: 用于根据设定特征从所述 SNS网络成员资料数 据库选取关注人群以构建关注人群数据库;
关系圏数据库, 用于存储关系圏成员的特征信息、 联系人信息和关系圏成 员所处的层次。
优选地, 所述装置进一步包括:
关系圏显示模块, 用于按照层次来显示关系圏成员。
优选地, 所述装置进一步包括:
影响力计算模块: 用于选取所述关系圏中各个关系圏成员的关系信息和 /或 特征信息, 并根据所述关系信息和 /或特征信息计算所述各个关系圏成员在所述 关系圏中的影响力。
本发明通过对目标人群的关系链进行分析, 根据特定的特征过滤条件提取 所述目标人群的关系圏成员并构建关系圏, 从而提供了 SNS网络中满足指定特 征的各个网络成员以及他们之间的关系。 利用目标人群的关系圏, 能够找到有 价值的关系链信息, 实现信息的精确搜索和传递, 方便商业活动推广和合作。 并且本发明还可根据关系圏中各个关系圏成员的关系信息和 /或特征信息计算所 述各个关系圏成员在所述关系圏中的影响力, 有助于寻找到关系圏中最具有影 响力的人群 , 使得信息的传递和检索更具有针对性。 附图说明
下面将结合附图及实施例对本发明作进一步说明, 附图中:
图 1是本发明的一种 SNS网络中成员关系圏的提取方法的第一实施例的流 程图;
图 2是本发明的一种 SNS网络中成员关系圏的提取方法的第二实施例的流 程图;
图 3是本发明的一种 SNS网络中成员关系圏的提取方法的目标人员的关系 圏成员及其层次的判定流程图;
图 4是本发明的一种 SNS网络中成员关系圏的提取方法的第三实施例的流 程图;
图 5是本发明的一种 SNS网络中成员关系圏的提取方法的影响力计算的步 骤流程图;
图 6是本发明的一种 SNS网络中成员关系圏的提取装置的第一实施例的结 构示意图;
图 7是本发明的一种 SNS网络中成员关系圏的提取装置的第二实施例的结 构示意图;
图 8是使用本发明的方法或装置的实施例提取的特定人群的关系圏的示意 图;
图 9是本发明的目标人群的关系圏的提取过程的示意图。 具体实施方式
为了使本发明的目的、 技术方案及优点更加清楚明白, 以下结合附图及实 施例, 对本发明进行进一步详细说明。 应当理解, 此处所描述的具体实施例仅 仅用以解释本发明, 并不用于限定本发明。
在本发明中, 可通过对目标人群的关系链进行分析, 根据特定的特征过滤 条件形成以目标人群作为关系链起点的关系圏。 网络用户或经销商可利用目标 人群的关系圏, 能够找到有价值的关系链信息, 实现信息的精确搜索和传递, 方便商业活动推广和合作。
图 1是本发明的一种 SNS网络中成员关系圏的提取方法的第一实施例的流 程图, 其具体过程如下:
在步骤 S101中,首先在 SNS网络中选择目标人群。 目标人群的选取可以根 据特定的查询条件从 SNS网络成员中选择, 比如查询条件可以是体育或者娱乐 圏的明星。 也可以是某个或是某些指定的网络成员, 比如, 可直接指定贝克汉 姆、 罗纳尔多和罗纳尔.迪尼奥为目标人群。
在步骤 S102中, 分析所述目标人群的关系链, 并根据特征过滤条件从所述 目标人群的关系链中提取所述目标人群的关系圏。 在一个实施例中, 所述关系 链可以包括目标人群的所有联系人信息, 该联系人与其他 SNS网络成员间的联 系信息以及其他相关信息。 比如, 目标人物 A认识 B , B认识 C , 那么 A和 B 有直接联系, 而 A与 C有间接联系。 A、 B和 C以及他们的相关信息就构成了 目标人物的一个最筒关系链。
图 2是本发明的一种 SNS网络中成员关系圏的提取方法的第二实施例的流 程图, 其具体过程如下:
在步骤 S201中,可以在 SNS网络中的服务器端建立关注人群数据库,并根 据查询条件从关注人群数据库中选取目标人群。 该关注人群数据库可以是从现 实世界或是媒体收集而来的, 这里可以是通过人工的方式或者是门户网站的新 闻统计装置, 也可以是根据搜索引擎的搜索排名获得的。
并且, 在本发明的一个实施例中, 关注人群数据库的建立可采用下列方式 来实现。 首先构建用于存放 SNS网络成员资料的 SNS网络成员数据库。 然后采 用自行开发或是第三方设计的特征统计规则数据库在的 SNS网络成员数据库中 提取特征关键词。 然后根据上述特征关键词从所述 SNS网络成员资料数据库提 取网络成员特征。 最后按照需要, 根据设定特征从所述 SNS网络成员资料数据 库中选取关注人群以构建关注人群数据库。 在该关注人群数据库中, 可存储有 所述关注人群的特征信息和所述关注人群对应于 SNS网络的唯一身份标识。 每 个关注人的特征信息可表示成(类型, 值), 比如用户 A最喜爱的电影是美丽心 灵, 描述为: (最喜爱的电影, 美丽心灵)。 而其对应于 SNS网络的唯一身份标识 可以是即时消息 (Instant message, IM)的登陆帐号。表 1示出了关注人在关注人群 数据库中存储的信息。
表 1 关注人群信息
Figure imgf000009_0001
而目标人群的选取, 可以是基于关注人群数据库的。 比如可以设定查询条 件为世界级的球星, 在关注人群数据库中选取满足该条件的关注人群, 如贝克 汉姆、 罗纳尔多等为目标人群。
在本发明的筒化实施例中,可以直接设定较为复杂的查询条件,直接在 SNS 网络成员数据库中选取目标人群。 在本发明的又一筒化实施例中, 可以直接指 定目标人群。
在步骤 S202中, 可设定目标人群的关系圏的规模预定值、 层次预定值和关 系圏成员的特征过滤条件。 比如, 前面选择贝克汉姆、 罗纳尔多等为目标人群, 可设定他们的关系圏的总人数为 900人, 层次为 3层, 每一层的人数分别为 300 人(每一层的人数可以相同,可以不同)。特征过滤信息可以是 (职业,球员),(其 他, 拍过广告)。 因此, 在该实施例中, 需要找出目标人群的关系圏中拍过广告 并且职业为球员的人群。 在步骤 S203中,从所述目标人群的关系链获取所述目标人群的联系人信息。 所述联系人信息可以是联系人对应于 SNS网络的唯一身份标识、 特征信息、 关 系类型和关系权重。 该联系人信息可以包括目标人群利用 SNS网络直接联系或 间接联系的联系人列表。 比如 ΙΜ好友列表、 blog的访问用户等等。 并且该目标 人群与联系人的关系可表示为(ID, type , value)。 ID表示联系人在 SNS网络中的 唯一身份标识, type表示关系的类型, 例如定义为: 好友, 认识, 陌生人。 value 定义为关系的权重, 即关系的重要程度。 权重越大表示关系越好, 联系越紧密, 越频繁。 本领域技术人员可以采用任何已知的方法或资料来确定目标人群的联 系人信息。
在步骤 S204中, 根据所述特征过滤条件和联系人信息选取所述目标人员的 关系圏成员并判定所述关系圏成员在所述关系圏中所处的层次。
在步骤 S205中, 根据关系圏数据库, 按照层次显示所述关系圏成员。 在本 发明的一个实施例中, 所述显示内容包括: 关系圏成员的特征信息、 关系类型、 权重信息和关系路径。
图 9示出了依照图 2示出的方法提取本发明的目标人群的关系圏的提取过 程。 图 8 示出了一个多层关系圏。 图中用线的不同风格表示关系的类型, 并且 进一步用线上数值表示关系实际权重, 权重越大, 关系越紧密。 每一环代表关 系的一个层次。 每两个关系圏成员存在直接或间接的关系, 关系的边连接起来 即为路经。
图 3示出了图 2中步骤 204的一个典型实施例。 如图 3所示, 在步骤 S301 中, 将关系圏的当前层次设定为 0, 并将所述目标人群作为关系圏成员添加到未 处理队列中, 同时清空已处理队列。
在步骤 S302中, 判断当前关系圏的规模是否达到预定值, 如果是则完成关 系圏提取步骤, 流程结束; 否则执行步骤 S303。 在该步骤中, 对关系圏规模的 判断可以包括对每一层关系圏成员人数的判断以及总关系圏人数的判断。
在步骤 S303中, 判断未处理队列是否为空, 如果是则完成关系圏提取步骤, 流程结束, 否则执行步骤 S304。
在步骤 S304中, 从未处理队列移走第一个关系圏成员作为当前关系圏成员 并将其添加到已处理队列, 并将当前层次设置为所述当前关系圏成员的层次。 在步骤 S305中, 提取所述当前关系圏成员的特征信息, 并判定当前层次是 否为 0, 如果是则执行步骤 S307 , 否则执行步骤 S306;
在步骤 S306中, 判定所述当前关系圏成员的特征信息是否满足特征过滤条 件, 如果是, 则执行步骤 S307, 否则执行步骤 S302;
在步骤 S307中, 判断当前层次 + 1是否小于预定层次, 如果是则执行步骤 S308 , 否则执行步骤 S302;
在步骤 S308中, 遍历当前关系圏成员的每一个联系人, 当所述联系人既不 在未处理队列也不在已处理队列时执行步骤 S309 , 否则执行步骤 S310;
在步骤 S309中, 将所述联系人作为关系圏成员添加到未处理队列的尾部, 并将所述联系人的层次设置为当前层次 + 1 , 再执行步骤 310;
在步骤 S310中, 将当前关系圏成员的特征信息、 联系人信息和层次存储到 关系圏数据库中, 再执行步骤 S302。 在该步骤中, 该关系圏成员的特征信息、 联系人信息和层次的存储可参见表 2。
表 2关系圈成员信息
Figure imgf000011_0001
图 4是本发明的一种 SNS网络中成员关系圏的提取方法的第三实施例的流 程图, 其具体过程如下:
在步骤 S401中, 首先在 SNS网络中, 选择目标人群。 目标人群的选取可以 根据特定的查询条件从 SNS网络成员中选择, 比如查询条件可以是体育或者娱 乐圏的明星。 也可以是某个或是某些指定的网络成员, 比如, 可直接指定贝克 汉姆、 罗纳尔多和罗纳尔.迪尼奥为目标人群
在步骤 S402中, 分析所述目标人群的关系链, 并根据特征过滤条件从所述 目标人群的关系链中提取所述目标人群的关系圏。
在步骤 S403中,选取所述关系圏中各个关系圏成员的关系信息和 /或特征信 息, 并根据所述关系信息和 /或特征信息计算所述各个关系圏成员在所述关系圏 中的影响力。 其中, 所述关系信息和 /或特征信息可以直接从所述目标人群的关 系链中获取; 也可从关注人群数据库中获取。 其中关注人群数据库的建立可以 参照前文的描述。 关系圏成员的特征信息可表示成(类型, 值), 比如用户 A最 喜爱的电影是美丽心灵, 描述为: (最喜爱的电影, 美丽心灵)。
根据本发明的方法, 可根据计算所述各个关系圏成员在所述关系圏中的影 响力, 有助于寻找到关系圏中最具有影响力的人群, 使得信息的传递和检索更 具有针对性。
图 5是图 4中影响力计算步骤的一个典型实施例。 如图 5所示,
在步骤 S501中, 选定一个关系圏, 并提取所述关系圏的特征过滤条件。 在步骤 S502中, 遍历所述关系圏中的每个成员, 选取每个关系圏成员的特 征信息, 并根据所述关系圏成员的特征信息和所述关系圏的特征过滤信息之间 的匹配程度计算所述关系圏成员的特征评分。
对于属于某一关系圏中的各个关系圏成员的特征评分可如下计算:
S (关系圏成员 ID) = {该关系圏成员的特征信息和所述关系圏的特征过滤 信息之间的匹配程度 }。
例如, 提取特征过滤信息为玩网络游戏的关系圏时, 可以为该用户玩该网 络游戏的信息, 包括时间, 等级的情况来转换相应的游戏积分, 从而作为特征 评分。
在步骤 S503中, 遍历所述关系圏中的每个成员, 选取关系圏成员的关系信 息, 并根据所述关系圏成员的关系信息计算所述关系圏成员的关系评分。 其中 所述关系信息包括关系圏成员间的关系类型、 权重信息和关系路径。
在本发明的一个实施例中, 特征评分函数设计如下:
R (关系圏成员 ID ) = {与其他每个关系圏成员的关系类型, 如果是好友 加 10分, 认识加 5分, 陌生加 1分}
在本发明的另一个实施例中, 特征评分函数设计如下:
R (关系圏成员 ID ) = {∑与其他每个关系圏成员的关系类型 *权重 }。 其 中, 权重表示关系的重要程度。 权重越大表示关系越好, 联系越紧密, 越频繁。 例如, 当关系类型为好友时计 10分, 当关系类型为认识时计 5分, 当关系类型 为陌生时计 1分。 当一个月内联系次数超过 5次时计 3分,联系次数为 3-5次时 计 2分, 少于 3次时计 1分。
在步骤 S504中, 根据所述特征评分和关系评分的加权结果计算所述关系圏 成员的影响力。
在本发明的一个实施例中, 影响力评分函数设计如下:
E (关系圏成员 10) =特征评分 *f+关系评分 *(l-f).
特征评分越大表明成员与关系圏的特征越吻合, 其影响力就越大。
关系评分越大表明成员与关系圏其他成员的关系越紧密, 其影响力就越大。 f为影响力权重, 默认为 0.5, 可根据实际需要调整。
在本发明的一个筒化实施例中, 可直接根据特征信息来计算关系圏成员的 影响力, 下面将其称为特征影响力。 其过程如下:
1、 选定一个关系圏, 并提取所述关系圏的特征过滤条件。
2、 遍历所述关系圏中的每个成员, 选取每个关系圏成员的特征信息, 并根 据所述关系圏成员的特征信息和所述关系圏的特征过滤信息之间的匹配程度计 算所述关系圏成员的特征影响力。
对于属于某一关系圏中的各个关系圏成员的特征影响力可如下计算:
SE (关系圏成员 ID) = {该关系圏成员的特征信息和所述关系圏的特征过滤 信息之间的匹配程度 }。
关系圏成员的特征影响力越大, 证明该关系圏成员与关系圏的特征越吻合。 在本发明的又一筒化实施例中, 可直接根据关系信息来计算关系圏成员的 影响力, 下面将其称为关系影响力。 其过程如下:
1、 选定一个关系圏, 遍历所述关系圏中的每个成员, 选取关系圏成员的关 系信息。 其中所述关系信息包括关系圏成员间的关系类型、 权重信息和关系路 径。
2、 ^^据所述关系圏成员的关系信息计算所述关系圏成员的关系影响力。 对于属于某一关系圏中的各个关系圏成员的关系影响力可如下计算:
RE (关系圏成员 ID ) = {∑与其他每个关系圏成员的关系类型 *权重 }。 或 RE (关系圏成员 ID ) = {与其他每个关系圏成员的关系类型, 如果是 好友加 10分, 认识加 5分, 陌生加 1分} 关系圏成员的特征关系越大, 证明其与关系圏其他成员的关系越紧密。 图 6是本发明的一种 SNS网络中成员关系圏的提取装置的第一实施例的结 构示意图。 该实施例为本发明的一个最筒实施例。 本发明的装置包括: 目标人 群选取模块 401和关系圏提取模块 402。所述目标人群选取模块 401可用于选择 目标人群。 目标人群的选取可以根据特定的查询条件从 SNS网络成员中选择, 也可以由客户直接指定某一 SNS网络成员。 比如查询条件可以是体育或者娱乐 圏的明星, 也可直接指定贝克汉姆、 罗纳尔多和罗纳尔.迪尼奥为目标人群。
所述关系圏提取模块 402, 用于分析所述目标人群的关系链, 并根据特征过 滤条件从所述目标人群的关系链中提取所述目标人群的关系圏。 在一个实施例 中, 所述关系链可以包括目标人群的所有联系人信息, 该联系人与其他 SNS网 络成员间的联系信息以及其他相关信息。 比如, 目标人物 A认识 B , B认识 C , 那么 A和 B有直接联系, 而 A与 C有间接联系。 A、 B和 C以及他们的相关信 息就构成了目标人物的一个最筒关系链。
在本发明的又一实施例中, 本发明的一种 SNS网络中成员关系圏的提取装 置还包括影响力计算模块(未示出)。 所述影响力计算模块可用于根据关系圏中 各个关系圏成员的关系信息和 /或特征信息计算所述各个关系圏成员在所述关系 圏中的影响力。 其中, 所述关系信息和 /或特征信息可以直接从所述目标人群的 关系链中获取; 也可从关注人群数据库中获取。 其中关注人群数据库的建立可 以参照前文的描述。
图 7是本发明的一种 SNS网络中成员关系圏的提取装置的第二实施例的结 构示意图。 在该实施例中, SNS网络成员资料数据库 501用于存放 SNS网络成 员的资料。 该资料可以是网络成员注册的时候自行填写的, 也可以是搜索引擎 从网络媒体公开的资料中收集的。 该资料可包括: 网络成员对应 SNS网络的唯 一身份标识、 特征信息、 关系链等任何在网络上或是现实生活中出现过的或是 有过记载的信息。
特征规则数据库 502与所述 SNS网络成员资料数据库 501相连以用于存储 从所述 SNS网络成员资料数据库 501获取的特征关键词。 该特征关键词的提取 可以采用本领域中公开的任何方法和规则进行, 本领域技术人员熟悉并能够熟 练应用这些方法和规则。 特征提取模块 503通过所述特征规则数据库 502与所 述 SNS网络成员资料数据库 501相连, 以从所述特征规则数据库 502选取所述 的特征关键词, 并根据所述特征关键词从所述 SNS网络成员资料数据库 501中 提取网络成员特征。 关注人群数据库构建模块 514可与所述 SNS网络成员资料 数据库 501、特征提取模块 503和特征规则数据库 502连接以根据设定的特征从 所述 SNS网络成员资料数据库 501选取关注人群进入关注人群数据库 504。 在 本发明的其他实施例中, 也可采用其他的方式构建关注人群数据库 504。
在本实施例中, 目标人群选取模块 505可以关注人群数据库 504相连, 并 从中选择目标人群。 在本发明的其他实施例中, 所述目标人群选取模块 505也 可直接接收外部输入, 或采用其他方式获取目标人群。
在本实施例中, 关系圏提取模块 506包括关系获取单元 5061和关系圏构建 单元 5062。在本发明的一个实施例中, 所述关系获取单元 5061可同时与目标人 群选取模块 505和 SNS网络成员资料数据库 501相连, 并从目标选取模块 505 获取目标人群, 接着根据所述目标人群的对应 SNS网络的唯一身份标识从 SNS 网络成员资料数据库 501中获取所述目标人群的关系链从而获得联系人信息。
所述联系人信息可以是联系人对应于 SNS网络的唯一身份标识、特征信息、 关系类型和关系权重。 该联系人信息可以包括目标人群利用 SNS网络直接联系 或间接联系的联系人列表。 比如 IM好友列表、 blog的访问用户等等。 并且该目 标人群与联系人的关系可表示为 (ID, type , value)。 ID表示联系人在 SNS网络中 的唯一身份标识, type表示关系的类型, 例如定义为: 好友, 认识, 陌生人。 value定义为关系的权重, 即关系的重要程度。 权重越大表示关系越好, 联系越 紧密, 越频繁。
在本发明的另一实施例中, 所述目标人群选取模块 505在选取目标人群的 同时, 获取关系链。 所述关系获取单元 5061可直接从目标人群选取模块 505获 取联系人信息。
在本发明的其他实施例中, 所述关系获取单元 5061还可采用其他本领域中 已知的技术从其他模块(如关注人群数据库 504 )获取目标人群的关系链从而得 到联系人信息。
关系圏构建单元 5062可根据特征过滤条件和联系人信息选取所述目标人员 的关系圏成员并判定所述关系圏成员在所述关系圏中所处的层次。 关系圏成员 的选取和层次设定可参见图 3。 本领域技术人员也可根据本发明的教导, 采用其 他的判断流程来完成该步骤。
在本发明的一个实施例中, 所述关系圏显示模块 508可与所述关系圏构建 单元 5062直接相连, 以按照层次来显示关系圏成员 (未示出)。
在本发明的另一实施例中, 所述关系圏构建单元 5062构建的关系圏可存储 到关系圏数据库 507中。 所述关系圏显示模块 508可与关系圏数据库 507相连 (参见图 7 ), 进而按照层次显示关系圏成员。 关系圏数据库 507中存储的每个 关系圏数据均是采用表格形式, 表格中可包括每个成员的 ID、 特征信息、 联系 人信息和层次信息。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程, 是可以通过计算机程序来指令相关的硬件来完成, 所述的程序可存储于一计算 机可读取存储介质中, 该程序在执行时, 可包括如上述各方法的实施例的流程。 其中, 所述的存储介质可为磁碟、 光盘、 只读存储记忆体(Read-Only Memory, ROM )或随机存储记忆体(Random Access Memory, RAM )等。
以上所述仅为本发明的较佳实施例而已, 并不用以限制本发明, 凡在本发 明的精神和原则之内所作的任何修改、 等同替换和改进等, 均应包含在本发明 的保护范围之内。

Claims

权利要求
1、 一种社会性网络服务 SNS网络中成员关系圏的提取方法, 其特征在于, 包括以下步骤:
(a)在所述 SNS网络中选择目标人群;
(b) 分析所述目标人群的关系链,并根据特征过滤条件从所述目标人群的关 系链中提取所述目标人群的关系圏。
2、根据权利要求 1所述的一种 SNS网络中成员关系圏的提取方法,其特征 在于, 所述步骤 (a)包括:
(al) 建立关注人群数据库, 并根据查询条件从所述关注人群数据库中选取 目标人群;
(a2)设定所述目标人群的关系圏的规模预定值、 层次预定值和关系圏成员 的特征过滤条件。
3、根据权利要求 2所述的一种 SNS网络中成员关系圏的提取方法,其特征 在于, 所述关注人群数据库中存储有所述关注人群的特征信息和所述关注人群 对应于 SNS网络的唯一身份标识。
4、根据权利要求 2所述的一种 SNS网络中成员关系圏的提取方法,其特征 在于, 所述步骤 (b)中进一步包括:
(bl)从所述目标人群的关系链获取所述目标人群的联系人信息;
(b2)根据所述特征过滤条件和联系人信息选取所述目标人群的关系圏成员 并判定所述关系圏成员在所述关系圏中所处的层次。
5、根据权利要求 4所述的一种 SNS网络中成员关系圏的提取方法,其特征 在于, 所述步骤 (b2)中进一步包括:
(b21) 将关系圏的当前层次设定为 0, 并将所述目标人群作为关系圏成员添 加到未处理队列中, 同时清空已处理队列;
(b22) 判断当前关系圏的规模是否达到规模预定值,如果是则完成关系圏提 取步骤, 否则执行步骤 (b23);
(b23) 判断未处理队列是否为空, 如果是则完成关系圏提取步骤, 否则从未 处理队列移走第一个关系圏成员作为当前关系圏成员并将其添加到已处理队 列, 并将当前层次设置为所述当前关系圏成员的层次;
(b24)提取所述当前关系圏成员的特征信息; 并判定当前层次是否为 0, 如 果是则执行步骤 (b26), 否则执行步骤 (b25);
(b25) 判定所述当前关系圏成员的特征信息是否满足特征过滤条件, 如果 是, 则执行步骤 (b26), 否则执行步骤 (b22) ;
(b26) 判断当前层次 + 1是否小于预定层次, 如果是则执行步骤 (b27), 否则 执行步骤 (b22);
(b27)遍历当前关系圏成员的每一个联系人, 当所述联系人既不在未处理队 列也不在已处理队列时将所述联系人作为关系圏成员添加到未处理队列的尾 部, 并将所述联系人的层次设置为当前层次 + 1 , 再执行步骤 (b28); 否则直接执 行步骤 b(28);
(b28) 将当前关系圏成员的特征信息、联系人信息和层次存储到关系圏数据 库中, 再执行步骤 b(22)。
6、根据权利要求 1或 5所述的一种 SNS网络中成员关系圏的提取方法, 其 特征在于, 所述方法进一步包括:
(c)根据所述关系圏数据库中的信息按照层次显示所述关系圏成员。
7、根据权利要求 1或 5所述的一种 SNS网络中成员关系圏的提取方法, 其 特征在于, 所述方法进一步包括:
(d) 选取所述关系圏中各个关系圏成员的关系信息和 /或特征信息, 并根据 所述关系信息和 /或特征信息计算所述各个关系圏成员在所述关系圏中的影响 力。
8、根据权利要求 7所述的一种 SNS网络中成员关系圏的提取方法,其特征 在于, 所述步骤 (d)进一步包括: (dl) 选取关系圏成员的特征信息, 并根据所述关系圏成员的特征信息与所 述关系圏的特征过滤信息之间的匹配程度计算所述关系圏成员的特征评分;
(d2) 选取关系圏成员的关系信息, 并根据所述关系圏成员的关系信息计算 所述关系圏成员的关系评分;
(d3)根据所述特征评分和关系评分的加权结果计算所述关系圏成员的影响 力。
9、 一种社会性网络服务 SNS网络中成员关系圏的提取装置, 其特征在于, 所述装置包括: 目标人群选取模块, 用于在所述 SNS网络中选择目标人群; 关系圏提取模块, 用于分析所述目标人群的关系链, 并根据特征过滤条件 从所述目标人群的关系链中提取所述目标人群的关系圏。
10、 根据权利要求 9所述的装置, 其特征在于, 所述关系圏提取模块包括: 关系获取单元: 用于从所述目标人群的关系链获取所述目标人群的联系人 信息; 关系圏构建单元: 用于根据所述特征过滤条件和联系人信息选取所述目标 人群的关系圏成员并判定所述关系圏成员在所述关系圏中所处的层次。
11、根据权利要求 9或 10所述的装置, 其特征在于, 所述装置进一步包括: 关注人群数据库, 用于存储所述关注人群的特征信息和所述关注人群对应 于 SNS网络的唯一身份标识; 并且 所述目标人群获取模块根据查询条件从所述关注人群数据库中选取目标人 群。
12、 根据权利要求 11所述的装置, 其特征在于, 所述装置进一步包括:
SNS网络成员资料数据库: 用于存放 SNS网络成员的资料; 特征规则数据库: 用于存储从所述 SNS网络成员资料数据库获取的特征关 键词; 特征提取模块: 用于根据所述特征关键词从所述 SNS网络成员资料数据库 提取网络成员特征; 关注人群数据库构建模块: 用于根据设定特征从所述 SNS网络成员资料数 据库选取关注人群以构建关注人群数据库; 关系圏数据库, 用于存储关系圏成员的特征信息、 联系人信息和关系圏成 员所处的层次。
13、 根据权利要求 12所述的装置, 其特征在于, 所述装置进一步包括: 关系圏显示模块, 用于根据所述关系圏数据库中的信息按照层次来显示关 系圏成员。
14、根据权利要求 9或 10所述的装置, 其特征在于, 所述装置进一步包括: 影响力计算模块: 用于根据关系圏中各个关系圏成员的关系信息和 /或特征 信息计算所述各个关系圏成员在所述关系圏中的影响力。
PCT/CN2010/070309 2009-03-12 2010-01-21 一种sns网络中成员关系圈的提取方法和装置 WO2010102527A1 (zh)

Priority Applications (5)

Application Number Priority Date Filing Date Title
BRPI1009430A BRPI1009430A2 (pt) 2009-03-12 2010-01-20 método e dispositivo para extração de círculos de relacionamentos de membros em redes de serviços de redes sociais (sns)
MX2011009462A MX2011009462A (es) 2009-03-12 2010-01-21 Metodo y dispositivo para extraer un circulo de relaciones de los miembros en una red de servicio de red social.
RU2011140609/08A RU2011140609A (ru) 2009-03-12 2010-01-21 Способ и устройство для извлечения круга взаимосвязей участников в сети службы социальной сети (ссс (sns)
CA2753096A CA2753096A1 (en) 2009-03-12 2010-01-21 Method and device for extracting relationship circle of members in social network service (sns) network
US13/230,269 US9031972B2 (en) 2009-03-12 2011-09-12 Method and device for extracting relationship circle of members in social network service (SNS) network

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CNA2009101059783A CN101510856A (zh) 2009-03-12 2009-03-12 一种sns网络中成员关系圈的提取方法和装置
CN200910105978.3 2009-03-12

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US13/230,269 Continuation US9031972B2 (en) 2009-03-12 2011-09-12 Method and device for extracting relationship circle of members in social network service (SNS) network

Publications (1)

Publication Number Publication Date
WO2010102527A1 true WO2010102527A1 (zh) 2010-09-16

Family

ID=41003122

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2010/070309 WO2010102527A1 (zh) 2009-03-12 2010-01-21 一种sns网络中成员关系圈的提取方法和装置

Country Status (7)

Country Link
US (1) US9031972B2 (zh)
CN (1) CN101510856A (zh)
BR (1) BRPI1009430A2 (zh)
CA (1) CA2753096A1 (zh)
MX (1) MX2011009462A (zh)
RU (1) RU2011140609A (zh)
WO (1) WO2010102527A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110569413A (zh) * 2019-08-16 2019-12-13 湖北工业大学 一种社交网络中保护成员隐私的索引系统及方法

Families Citing this family (22)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101510856A (zh) 2009-03-12 2009-08-19 腾讯科技(深圳)有限公司 一种sns网络中成员关系圈的提取方法和装置
CN101820452B (zh) * 2009-12-14 2013-08-28 宇龙计算机通信科技(深圳)有限公司 一种获取联系人相关信息的方法及移动终端
US20110173316A1 (en) * 2010-01-13 2011-07-14 c/o Microsoft Corporation Relationship based representation of participants in shared online space
US8874773B2 (en) 2010-11-30 2014-10-28 Gary W. Grube Obtaining group and individual emergency preparedness communication information
CN102724168B (zh) * 2011-03-30 2014-10-01 腾讯科技(深圳)有限公司 一种推送用户信息的方法及装置
CN102833176B (zh) * 2011-06-13 2018-01-26 腾讯科技(深圳)有限公司 获取信息的方法、装置和系统
JP5866922B2 (ja) * 2011-09-22 2016-02-24 富士ゼロックス株式会社 検索装置及びプログラム
CN103136303A (zh) * 2011-11-24 2013-06-05 北京千橡网景科技发展有限公司 在社交网络服务网站中划分用户群组的方法和设备
CN103164416B (zh) * 2011-12-12 2016-08-03 阿里巴巴集团控股有限公司 一种用户关系的识别方法及设备
US9721039B2 (en) * 2011-12-16 2017-08-01 Palo Alto Research Center Incorporated Generating a relationship visualization for nonhomogeneous entities
CN103425648B (zh) * 2012-05-15 2016-04-13 腾讯科技(深圳)有限公司 关系圈的处理方法和系统
CN103488651B (zh) * 2012-06-08 2018-12-11 北京千橡网景科技发展有限公司 用于获得成员之间的关系路径的方法和装置
CN104348871B (zh) * 2013-08-05 2019-01-11 深圳市腾讯计算机系统有限公司 一种同类账号扩展方法及装置
WO2016037277A1 (en) * 2014-09-10 2016-03-17 Sysomos L.P. Systems and methods for identifying a target audience in a social data network
CN105225149B (zh) * 2015-09-07 2018-04-27 腾讯科技(深圳)有限公司 一种征信评分确定方法及装置
CN105554140B (zh) * 2015-12-24 2020-01-10 腾讯科技(深圳)有限公司 一种用户群体定位方法及服务器
CN106327326A (zh) * 2016-08-30 2017-01-11 北京小米移动软件有限公司 信用评级的方法及装置
US10606866B1 (en) 2017-03-30 2020-03-31 Palantir Technologies Inc. Framework for exposing network activities
CN107688605B (zh) 2017-07-26 2019-02-26 平安科技(深圳)有限公司 跨平台数据匹配方法、装置、计算机设备和存储介质
CN109408543B (zh) * 2018-09-26 2021-07-23 北京华宝智慧科技有限公司 一种智能关系网嗅探方法
CN111275563A (zh) * 2020-01-21 2020-06-12 深圳市比量科技传媒有限公司 基于微信动作的人脉关系的生成方法、系统及存储介质
CN116226460A (zh) * 2022-12-09 2023-06-06 中科世通亨奇(北京)科技有限公司 基于人物图谱的最具价值路径提取方法及设备

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0924919A2 (en) * 1997-12-18 1999-06-23 Nortel Networks Corporation System and method for centrex translation
JP2007328723A (ja) * 2006-06-09 2007-12-20 Fujifilm Corp 情報提供装置および情報提供方法
CN101197878A (zh) * 2006-12-09 2008-06-11 国际商业机器公司 用于通信转发的方法和系统
JP2008146246A (ja) * 2006-12-07 2008-06-26 Taito Corp Snsサーバ、sns制御プログラム
CN101510856A (zh) * 2009-03-12 2009-08-19 腾讯科技(深圳)有限公司 一种sns网络中成员关系圈的提取方法和装置

Family Cites Families (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050060417A1 (en) * 2003-09-16 2005-03-17 Rose Keith R. Automated electronic personal preference & proxy network
US7421454B2 (en) * 2004-02-27 2008-09-02 Yahoo! Inc. Method and system for managing digital content including streaming media
US20060047419A1 (en) * 2004-09-02 2006-03-02 Diendorf John R Telematic method and apparatus for managing shipping logistics
US20070011665A1 (en) * 2005-06-21 2007-01-11 Microsoft Corporation Content syndication platform
US20070060129A1 (en) * 2005-09-14 2007-03-15 Jorey Ramer Mobile communication facility characteristic influenced search results
US7881969B2 (en) * 2005-12-13 2011-02-01 Microsoft Corporation Trust based architecture for listing service
US8027943B2 (en) * 2007-08-16 2011-09-27 Facebook, Inc. Systems and methods for observing responses to invitations by users in a web-based social network
US20090049127A1 (en) * 2007-08-16 2009-02-19 Yun-Fang Juan System and method for invitation targeting in a web-based social network
CA2634206A1 (en) 2005-12-22 2007-07-05 Catalog.Com, Inc. System and method for cross-domain social networking
US20070260568A1 (en) * 2006-04-21 2007-11-08 International Business Machines Corporation System and method of mining time-changing data streams using a dynamic rule classifier having low granularity
US7596597B2 (en) * 2006-08-31 2009-09-29 Microsoft Corporation Recommending contacts in a social network
US20080062244A1 (en) 2006-09-11 2008-03-13 Hewlett-Packard Development Company Lp Laser writing
US8108405B2 (en) * 2007-10-05 2012-01-31 Fujitsu Limited Refining a search space in response to user input
CN101216829B (zh) 2007-12-28 2010-06-02 丁景涛 内容搜索系统、设备及其方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0924919A2 (en) * 1997-12-18 1999-06-23 Nortel Networks Corporation System and method for centrex translation
JP2007328723A (ja) * 2006-06-09 2007-12-20 Fujifilm Corp 情報提供装置および情報提供方法
JP2008146246A (ja) * 2006-12-07 2008-06-26 Taito Corp Snsサーバ、sns制御プログラム
CN101197878A (zh) * 2006-12-09 2008-06-11 国际商业机器公司 用于通信转发的方法和系统
CN101510856A (zh) * 2009-03-12 2009-08-19 腾讯科技(深圳)有限公司 一种sns网络中成员关系圈的提取方法和装置

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110569413A (zh) * 2019-08-16 2019-12-13 湖北工业大学 一种社交网络中保护成员隐私的索引系统及方法
CN110569413B (zh) * 2019-08-16 2022-11-18 湖北工业大学 一种社交网络中保护成员隐私的索引系统及方法

Also Published As

Publication number Publication date
US20110320485A1 (en) 2011-12-29
CN101510856A (zh) 2009-08-19
BRPI1009430A2 (pt) 2016-03-01
RU2011140609A (ru) 2013-04-20
MX2011009462A (es) 2011-10-03
CA2753096A1 (en) 2010-09-16
US9031972B2 (en) 2015-05-12

Similar Documents

Publication Publication Date Title
WO2010102527A1 (zh) 一种sns网络中成员关系圈的提取方法和装置
KR101728122B1 (ko) 소셜 네트워크에서 사용자들을 추천하기 위한 방법 및 그 시스템
US11122009B2 (en) Systems and methods for identifying geographic locations of social media content collected over social networks
Craglia et al. Digital Earth from vision to practice: making sense of citizen-generated content
US10742574B2 (en) Method and device for implementing instant communication
CN104717124B (zh) 一种好友推荐方法、装置及服务器
US11310338B1 (en) Method and system for topic disambiguation and classification
CN106302085B (zh) 即时通讯群组的推荐方法及系统
WO2016119642A1 (zh) 一种信息处理方法、客户端及服务器、计算机存储介质
US20130297694A1 (en) Systems and methods for interactive presentation and analysis of social media content collection over social networks
US20130304818A1 (en) Systems and methods for discovery of related terms for social media content collection over social networks
US20130297581A1 (en) Systems and methods for customized filtering and analysis of social media content collected over social networks
KR20130139338A (ko) 얼굴 인식
WO2013026325A1 (zh) 一种人物搜索方法、装置及存储介质
Cornet et al. How image-based social media websites support social movements
US20090037315A1 (en) System and method for brokering agents and auditionees
CN111597446B (zh) 基于人工智能的内容推送方法、装置、服务器和存储介质
WO2018196553A1 (zh) 标识的获取方法及装置、存储介质以及电子装置
JP2010211569A (ja) 評価装置、プログラム及び情報処理システム
WO2014029314A1 (zh) 信息聚合归类的显示方法及系统
US11323404B2 (en) System and method for displaying message or user lists
CN111143689A (zh) 根据用户需求和用户画像构建推荐引擎的方法
US20190012601A1 (en) Information visualization method for user group decision making, and user terminal using said method, operation method of server providing information for user group decision making, and server apparatus using said method
JP7333931B2 (ja) 投稿分析システム、投稿分析装置および投稿分析方法
WO2021219117A1 (zh) 图像检索方法、图像检索装置、图像检索系统及图像显示系统

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 10750323

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 2753096

Country of ref document: CA

WWE Wipo information: entry into national phase

Ref document number: MX/A/2011/009462

Country of ref document: MX

NENP Non-entry into the national phase

Ref country code: DE

WWE Wipo information: entry into national phase

Ref document number: 6710/CHENP/2011

Country of ref document: IN

ENP Entry into the national phase

Ref document number: 2011140609

Country of ref document: RU

Kind code of ref document: A

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1)EPC

122 Ep: pct application non-entry in european phase

Ref document number: 10750323

Country of ref document: EP

Kind code of ref document: A1

REG Reference to national code

Ref country code: BR

Ref legal event code: B01A

Ref document number: PI1009430

Country of ref document: BR

ENP Entry into the national phase

Ref document number: PI1009430

Country of ref document: BR

Kind code of ref document: A2

Effective date: 20110909