WO2019205373A1 - 相似用户查找装置、方法及计算机可读存储介质 - Google Patents

相似用户查找装置、方法及计算机可读存储介质 Download PDF

Info

Publication number
WO2019205373A1
WO2019205373A1 PCT/CN2018/102112 CN2018102112W WO2019205373A1 WO 2019205373 A1 WO2019205373 A1 WO 2019205373A1 CN 2018102112 W CN2018102112 W CN 2018102112W WO 2019205373 A1 WO2019205373 A1 WO 2019205373A1
Authority
WO
WIPO (PCT)
Prior art keywords
node
graph
undirected
connectivity graph
nodes
Prior art date
Application number
PCT/CN2018/102112
Other languages
English (en)
French (fr)
Other versions
WO2019205373A9 (zh
Inventor
王健宗
吴天博
黄章成
肖京
Original Assignee
平安科技(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
Application filed by 平安科技(深圳)有限公司 filed Critical 平安科技(深圳)有限公司
Publication of WO2019205373A1 publication Critical patent/WO2019205373A1/zh
Publication of WO2019205373A9 publication Critical patent/WO2019205373A9/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90Details of database functions independent of the retrieved data types
    • G06F16/95Retrieval from the web
    • G06F16/953Querying, e.g. by the use of web search engines
    • G06F16/9536Search customisation based on social or collaborative filtering

Definitions

  • the present application relates to the field of computer technologies, and in particular, to a similar user search device, method, and computer readable storage medium.
  • the existing calculation method of user similarity mainly depends on user data.
  • This method has the following drawbacks: on the one hand, user data is difficult to reflect all the characteristics of a person, and on the other hand, if the user data is incomplete, the calculation of similarity is not ideal, therefore, calculation by this method The similar users obtained are less accurate.
  • the present application provides a similar user search device, method and computer readable storage medium, the main purpose of which is to improve the accuracy of finding similar users.
  • the present application provides a similar user search device based on community search, the device comprising a memory and a processor, wherein the memory stores a user finder running on the processor, the user searching
  • the program implements the following steps when executed by the processor:
  • the set of nodes in the generated sub-graph constitutes a community to which the target user belongs;
  • the user corresponding to the node in the generated subgraph is used as a similar user of the plurality of target users.
  • the present application further provides a similar user search method based on community search, the method includes:
  • the set of nodes in the generated sub-graph constitutes a community to which the target user belongs;
  • the user corresponding to the node in the generated subgraph is used as a similar user of the plurality of target users.
  • the present application further provides a computer readable storage medium having a user finder stored thereon, the user finder being executable by one or more processors to implement The following steps:
  • the set of nodes in the generated sub-graph constitutes a community to which the target user belongs;
  • the user corresponding to the node in the generated subgraph is used as a similar user of the plurality of target users.
  • the similar user searching device, method and computer readable storage medium construct an undirected connectivity graph according to user information in the target social network platform, wherein one node on the undirected connectivity graph corresponds to one user and has an association relationship
  • the users are connected by an edge to obtain a set of query nodes composed of a plurality of target users having an associated relationship.
  • the set of query nodes is a subset of the set of nodes in the spiced connected graph, and is preset according to the set of query nodes.
  • the community search algorithm performs an iterative operation on the undirected connected graph to delete the nodes on the undirected connected graph until the generated subgraphs satisfying the second preset condition are obtained after deleting the nodes, and the set of nodes in the generated subgraph constitutes a target
  • the community to which the user belongs, and on the undirected connectivity graph, the users belonging to one community have similar attributes. Therefore, the user corresponding to the node on the generated sub-picture can be used as the similar user of the target user.
  • the application searches the community through the social network platform, finds the community described by the target user, and obtains similar users of the target user, thereby improving the accuracy of finding similar users.
  • FIG. 1 is a schematic diagram of a preferred embodiment of a similar user search device based on community search in the present application
  • FIG. 2 is a schematic diagram of a program module of a user finder in an embodiment of a similar user search device based on community search;
  • FIG. 3 is a flowchart of a first embodiment of a similar user search method based on community search in the present application.
  • the application provides a similar user search device based on community search.
  • FIG. 1 a schematic diagram of a preferred embodiment of a similar user search device based on community search of the present application is shown.
  • the similar user search device 1 based on the community search may be a PC (Personal Computer), or may be a terminal device such as a smart phone, a tablet computer, or a portable computer.
  • the community search based similar user lookup device 1 includes at least a memory 11, a processor 12, a network interface 13, and a communication bus 14.
  • the memory 11 includes at least one type of readable storage medium including a flash memory, a hard disk, a multimedia card, a card type memory (for example, an SD or DX memory, etc.), a magnetic memory, a magnetic disk, an optical disk, and the like.
  • the memory 11 may be, in some embodiments, an internal storage unit of a similar user lookup device 1 based on a community search, such as a hard disk of the community search based similar user lookup device 1.
  • the memory 11 may also be an external storage device of a similar user search device 1 based on community search in other embodiments, such as a plug-in hard disk equipped on a similar user search device 1 based on community search, a smart memory card (Smart Media Card) , SMC), Secure Digital (SD) card, Flash Card, etc. Further, the memory 11 may also include an internal storage unit of the similar user search device 1 based on the community search and an external storage device.
  • the memory 11 can be used not only for storing application software and various types of data installed in the community search-like similar user search device 1, such as code of the user search program 01, but also for temporarily storing data that has been output or will be output.
  • the processor 12 may be a Central Processing Unit (CPU), controller, microcontroller, microprocessor or other data processing chip for running program code or processing stored in the memory 11. Data, such as executing user finder 01 and the like.
  • CPU Central Processing Unit
  • controller microcontroller
  • microprocessor or other data processing chip for running program code or processing stored in the memory 11.
  • Data such as executing user finder 01 and the like.
  • the network interface 13 can optionally include a standard wired interface, a wireless interface (such as a WI-FI interface), and is typically used to establish a communication connection between the device 1 and other electronic devices.
  • a standard wired interface such as a WI-FI interface
  • Communication bus 14 is used to implement connection communication between these components.
  • Figure 1 shows only a community search based similar user lookup device 1 with components 11-14 and user finder 01, but it should be understood that not all illustrated components may be implemented, alternative implementations may be Fewer components.
  • the device 1 may further include a user interface
  • the user interface may include a display
  • an input unit such as a keyboard
  • the optional user interface may further include a standard wired interface and a wireless interface.
  • the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch sensor, or the like.
  • the display may also be appropriately referred to as a display screen or a display unit for displaying information processed in the similar user search device 1 based on the community search and a user interface for displaying the visualization.
  • the user 11 is stored in the memory 11; when the processor 12 executes the user search program 01 stored in the memory 11, the following steps are implemented:
  • An undirected connectivity graph is constructed according to user information in the target social network platform, wherein, on the undirected connectivity graph, one node corresponds to one user, and users with associated relationships are connected by one edge.
  • the user information includes the user's follower information and the follower information, and the information reflects the relationship between the users.
  • the concern relationship may be the user's concern relationship, friend relationship, co-investment, etc. on the above platform.
  • the concern relationship can be one-way attention to the relationship or two-way attention to the relationship.
  • a complex network is constructed based on the relationship between users. On the complex network, one node corresponds to one user, and users with associated relationships are connected by one edge.
  • the social network platform includes a microblog platform, a Twitter platform, or a financial forum.
  • the user corresponding to the node in the generated subgraph is used as a similar user of the plurality of target users.
  • the set of query nodes formed by the target users, and the set of query nodes is a subset of the set of user nodes V of the graph G.
  • an iterative calculation is performed on the undirected connectivity graph G according to the community search algorithm to find a user group that has close relationship with these target users or has similar preferences.
  • These user groups and their associations form a generated subgraph of the undirected connectivity graph G.
  • the steps for generating a subgraph of a condition include:
  • the node in the node set has the minimum degree in the graph unconnected graph after the node delete operation, or the node in the query node set is no longer connected on the undirected connected graph after the node delete operation; if yes, the node is deleted. Operation, the current undirected connected graph is used as the generated sub-graph; if not, the node having the smallest degree in the deleted undirected connected graph and the edge connected to the node are continuously executed based on the current undirected connected graph A step of.
  • the cyclically deleted node having the smallest degree in each step of the graph until the node deletion operation is performed satisfies one of the following conditions: the node in the query node set has the minimum degree in the graph, or the deletion has the minimum After the nodes of degree, the points in the query node set are no longer connected.
  • the node deletion operation is terminated, and the generated sub-picture obtained after the calculation is terminated is obtained.
  • the generated sub-picture is a connected graph, and the query node set is a subset of the node set in the generated sub-picture.
  • the users corresponding to the nodes in the generated sub-graph obtained according to the above algorithm are the communities to be searched, and these users have the same features as the target users, for example, having the same product of interest or having the same feature tag.
  • the preset condition may be further increased: the maximum distance between the node in the generated subgraph and the node in the query node set Q does not exceed a preset distance. .
  • the undirected connectivity graph after the node deletion operation satisfies the second preset condition, before terminating the node deletion operation, calculating a distance between the node in the current undirected connectivity graph and the set of the query node, determining whether there is The distance from the node to the query node is greater than the preset distance; if yes, the node whose distance to the query node is greater than the preset distance is deleted, and the undirected connected graph after deleting the node is used as the generated sub-picture; if not, the execution will be performed The node deletion operation is terminated, and the current undirected connectivity graph is used as the step of generating the sub-picture.
  • the calculation of the distance between the nodes may be in any of the following ways: Mode 1, the number of edges in the shortest path between the two nodes is taken as the distance between the nodes due to a complex undirected connection.
  • the connection between two nodes can be realized by multiple paths, where the distance between the nodes can be calculated according to the shortest path among the multiple paths; the second method is based on the edge in the shortest path between the two nodes.
  • the weight calculation node calculates the distance between the nodes.
  • the user information obtained from the target social network platform includes tag information, and the weights between the two users are set according to the number of the same tags owned by the two users.
  • the tag information may be tags that the user can add to the social network platform, the tags may reflect the user's tag information of interest, or may also extract keywords that are extracted from the content posted by the user on the social network platform. As the user's tag information.
  • the distance from the node to each node in the query node set is first calculated, and then the minimum distance among the plurality of distances is used as the distance from the node to the query node set.
  • the community where the target user is located is obtained, that is, the community formed by the user corresponding to the node in the generated sub-picture, and the users are similar users of the target user, for example, in a financial forum, can be considered in the same community.
  • Users in the same account have the same or similar investment preferences, attributes, tags, events of interest, and so on. After getting similar users, make accurate recommendations or targeted delivery based on the same or similar investment preferences, attributes, tags, events of interest, etc. between them.
  • the similar user search device proposed in the above embodiment constructs an undirected connectivity graph according to the user information in the target social network platform, wherein one node on the undirected connectivity graph corresponds to one user, and the user with the associated relationship passes an edge Connected to obtain a set of query nodes composed of a plurality of target users having an associated relationship.
  • the set of query nodes is a subset of the set of nodes in the spiced connected graph, and the community search algorithm preset according to the set of query nodes is in the non-directional
  • the iterative operation is performed on the connected graph to delete the nodes on the undirected connected graph, and after the node is deleted, the generated sub-graph satisfying the second preset condition is obtained, and the set of nodes in the generated sub-graph constitutes the community to which the target user belongs, and
  • users belonging to one community have similar attributes. Therefore, the user corresponding to the node on the generated sub-picture can be used as a similar user of the target user.
  • the application searches the community through the social network platform, finds the community described by the target user, and obtains similar users of the target user, thereby improving the accuracy of finding similar users.
  • the user finder may also be divided into one or more modules, one or more modules being stored in the memory 11 and being processed by one or more processors (this embodiment is The processor 12) is executed to complete the application, and the module referred to in the present application refers to a series of computer program instruction segments capable of performing a specific function for describing the execution process of the user finder in a similar user search device based on community search. .
  • FIG. 2 it is a schematic diagram of a program module of a user finder in an embodiment of a similar user search device based on community search in the present application.
  • the user finder can be divided into an information generating module 10 and information.
  • the obtaining module 20, the community search module 30, and the user acquisition module 40 are exemplarily:
  • the information generating module 10 is configured to: construct an undirected connectivity graph according to user information in the target social network platform, where, on the undirected connectivity graph, one node corresponds to one user, and users with associated relationships pass one edge Connected
  • the information obtaining module 20 is configured to: acquire a set of query nodes formed by a plurality of target users having an associated relationship, where the set of query nodes is a subset of the set of nodes in the undirected connected graph;
  • the community search module 30 is configured to: perform an iterative operation on the undirected connectivity graph according to the query node set and a preset community search algorithm, to delete a node on the undirected connectivity graph, until the node is deleted Generating a sub-graph of the second preset condition, the set of nodes in the generated sub-graph constitutes a community to which the target user belongs;
  • the user obtaining module 40 is configured to: use a user corresponding to the node in the generated sub-picture as a similar user of the multiple target users.
  • the present application also provides a similar user search method based on community search.
  • FIG. 3 it is a flowchart of a first embodiment of a similar user search method based on community search of the present application. The method can be performed by a device that can be implemented by software and/or hardware.
  • a similar user search method based on community search includes:
  • Step S10 Construct an undirected connectivity graph according to user information in the target social network platform, wherein, on the undirected connectivity graph, one node corresponds to one user, and users with associated relationships are connected by one edge.
  • the user information includes the user's follower information and the follower information, and the information reflects the relationship between the users.
  • the concern relationship may be the user's concern relationship, friend relationship, co-investment, etc. on the above platform.
  • the concern relationship can be one-way attention to the relationship or two-way attention to the relationship.
  • a complex network is constructed based on the relationship between users. On the complex network, one node corresponds to one user, and users with associated relationships are connected by one edge.
  • the social network platform includes a microblog platform, a Twitter platform, or a financial forum.
  • Step S20 Acquire a set of query nodes composed of a plurality of target users having an association relationship, where the set of query nodes is a subset of the set of nodes in the undirected connectivity graph.
  • Step S30 Perform an iterative operation on the undirected connectivity graph according to the query node set and the preset community search algorithm, to delete the node on the undirected connectivity graph, and obtain the second preset after deleting the node.
  • a subgraph of the condition is generated, and the set of nodes in the generated subgraph constitutes a community to which the target user belongs.
  • Step S40 The user corresponding to the node in the generated subgraph is used as a similar user of the plurality of target users.
  • the set of query nodes formed by the target users, and the set of query nodes is a subset of the set of user nodes V of the graph G.
  • an iterative calculation is performed on the undirected connectivity graph G according to the community search algorithm to find a user group that has close relationship with these target users or has similar preferences.
  • These user groups and their associations form a generated subgraph of the undirected connectivity graph G.
  • step S30 includes:
  • the node in the node set has the minimum degree in the graph unconnected graph after the node delete operation, or the node in the query node set is no longer connected on the undirected connected graph after the node delete operation; if yes, the node is deleted. Operation, the current undirected connected graph is used as the generated sub-graph; if not, the node having the smallest degree in the deleted undirected connected graph and the edge connected to the node are continuously executed based on the current undirected connected graph A step of.
  • the cyclically deleted node having the smallest degree in each step of the graph until the node deletion operation is performed satisfies one of the following conditions: the node in the query node set has the minimum degree in the graph, or the deletion has the minimum After the nodes of degree, the points in the query node set are no longer connected.
  • the node deletion operation is terminated, and the generated sub-picture obtained after the calculation is terminated is obtained.
  • the generated sub-picture is a connected graph, and the query node set is a subset of the node set in the generated sub-picture.
  • the users corresponding to the nodes in the generated sub-graph obtained according to the above algorithm are the communities to be searched, and these users have the same features as the target users, for example, having the same product of interest or having the same feature tag.
  • the preset condition may be further increased: the maximum distance between the node in the generated subgraph and the node in the query node set Q does not exceed a preset distance. .
  • the undirected connectivity graph after the node deletion operation satisfies the second preset condition, before terminating the node deletion operation, calculating a distance between the node in the current undirected connectivity graph and the set of the query node, determining whether there is The distance from the node to the query node is greater than the preset distance; if yes, the node whose distance to the query node is greater than the preset distance is deleted, and the undirected connected graph after deleting the node is used as the generated sub-picture; if not, the execution will be performed The node deletion operation is terminated, and the current undirected connectivity graph is used as the step of generating the sub-picture.
  • the calculation of the distance between the nodes may be in any of the following ways: Mode 1, the number of edges in the shortest path between the two nodes is taken as the distance between the nodes due to a complex undirected connection.
  • the connection between two nodes can be realized by multiple paths, where the distance between the nodes can be calculated according to the shortest path among the multiple paths; the second method is based on the edge in the shortest path between the two nodes.
  • the weight calculation node calculates the distance between the nodes.
  • the user information obtained from the target social network platform includes tag information, and the weights between the two users are set according to the number of the same tags owned by the two users.
  • the tag information may be tags that the user can add to the social network platform, the tags may reflect the user's tag information of interest, or may also extract keywords that are extracted from the content posted by the user on the social network platform. As the user's tag information.
  • the distance from the node to each node in the query node set is first calculated, and then the minimum distance among the plurality of distances is used as the distance from the node to the set of query nodes.
  • the community where the target user is located is obtained, that is, the community formed by the user corresponding to the node in the generated sub-picture, and the users are similar users of the target user, for example, in a financial forum, can be considered in the same community.
  • Users in the same account have the same or similar investment preferences, attributes, tags, events of interest, and so on. After getting similar users, make accurate recommendations or targeted delivery based on the same or similar investment preferences, attributes, tags, events of interest, etc. between them.
  • the similar user search method proposed in the above embodiment constructs an undirected connectivity graph according to the user information in the target social network platform, wherein one node corresponds to one user and the user with the associated relationship passes an edge on the undirected connectivity graph Connected to obtain a set of query nodes composed of a plurality of target users having an associated relationship.
  • the set of query nodes is a subset of the set of nodes in the spiced connected graph, and the community search algorithm preset according to the set of query nodes is in the non-directional
  • the iterative operation is performed on the connected graph to delete the nodes on the undirected connected graph, and after the node is deleted, the generated sub-graph satisfying the second preset condition is obtained, and the set of nodes in the generated sub-graph constitutes the community to which the target user belongs, and
  • users belonging to one community have similar attributes. Therefore, the user corresponding to the node on the generated sub-picture can be used as a similar user of the target user.
  • the application searches the community through the social network platform, finds the community described by the target user, and obtains similar users of the target user, thereby improving the accuracy of finding similar users.
  • the embodiment of the present application further provides a computer readable storage medium, where the user finder is stored, and the user finder can be executed by one or more processors to implement the following operations:
  • the set of nodes in the generated sub-graph constitutes a community to which the target user belongs;
  • the user corresponding to the node in the generated subgraph is used as a similar user of the plurality of target users.
  • the specific embodiment of the computer readable storage medium of the present application is substantially the same as the embodiment of the similar user search device and method based on the community search, and is not described herein.
  • the technical solution of the present application which is essential or contributes to the prior art, may be embodied in the form of a software product stored in a storage medium (such as ROM/RAM as described above). , a disk, an optical disk, including a number of instructions for causing a terminal device (which may be a mobile phone, a computer, a server, or a network device, etc.) to perform the methods described in the various embodiments of the present application.
  • a terminal device which may be a mobile phone, a computer, a server, or a network device, etc.

Landscapes

  • Engineering & Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种基于社区搜索的相似用户查找装置,包括存储器和处理器,存储器上存储有可在处理器上运行的用户查找程序,该程序被处理器执行时实现如下步骤:根据目标社交网络平台中的用户信息构建无向连通图,其中;获取由多个具有关联关系的目标用户构成的查询节点集,查询节点集为无向连通图中的节点集合的子集;根据查询节点集和预设的社区搜索算法在无向连通图上进行迭代运算,以删除无向连通图上的节点,直至删除节点后得到满足第二预设条件的生成子图;将生成子图中的节点对应的用户作为多个目标用户的相似用户。所述装置提高了查找相似用户的准确度。

Description

相似用户查找装置、方法及计算机可读存储介质
本申请基于巴黎公约申明享有2018年04月26日递交的申请号为2018103823028、名称为“相似用户查找装置、方法及计算机可读存储介质”的中国专利申请的优先权,该中国专利申请的整体内容以参考的方式结合在本申请中。
技术领域
本申请涉及计算机技术领域,尤其涉及一种相似用户查找装置、方法及计算机可读存储介质。
背景技术
随着信息技术的发展,互联网、移动互联网、物联网能够收集到越来越多的用户信息,如何将这些信息采集、存储并分析,进而发现不同用户之间的相似性正在成为研究热点。
目前,现有的用户相似度的计算方法主要依赖于用户资料。这种方式存在以下缺陷:一方面,用户资料很难反映一个人的全部特征,而另一方面,如果用户资料填写不完整,那么相似度的计算是不理想的,因此,通过这种方法计算得到的相似用户的准确度较低。
发明内容
本申请提供一种相似用户查找装置、方法及计算机可读存储介质,其主要目的在于提高查找相似用户的准确度。
为实现上述目的,本申请提供一种基于社区搜索的相似用户查找装置,该装置包括存储器和处理器,所述存储器中存储有可在所述处理器上运行的用户查找程序,所述用户查找程序被所述处理器执行时实现如下步骤:
根据目标社交网络平台中的用户信息构建无向连通图,其中,在所述无向连通图上,一个节点对应于一个用户,具有关联关系的用户之间通过一条边相连接;
获取由多个具有关联关系的目标用户构成的查询节点集,所述查询节点 集为所述无向连通图中的节点集合的子集;
根据所述查询节点集和预设的社区搜索算法在所述无向连通图上进行迭代运算,以删除所述无向连通图上的节点,直至删除节点后得到满足第二预设条件的生成子图,该生成子图中的节点集合构成所述目标用户所属的社区;
将所述生成子图中的节点对应的用户作为所述多个目标用户的相似用户。
此外,为实现上述目的,本申请还提供一种基于社区搜索的相似用户查找方法,该方法包括:
根据目标社交网络平台中的用户信息构建无向连通图,其中,在所述无向连通图上,一个节点对应于一个用户,具有关联关系的用户之间通过一条边相连接;
获取由多个具有关联关系的目标用户构成的查询节点集,所述查询节点集为所述无向连通图中的节点集合的子集;
根据所述查询节点集和预设的社区搜索算法在所述无向连通图上进行迭代运算,以删除所述无向连通图上的节点,直至删除节点后得到满足第二预设条件的生成子图,该生成子图中的节点集合构成所述目标用户所属的社区;
将所述生成子图中的节点对应的用户作为所述多个目标用户的相似用户。
此外,为实现上述目的,本申请还提供一种计算机可读存储介质,所述计算机可读存储介质上存储有用户查找程序,所述用户查找程序可被一个或者多个处理器执行,以实现如下步骤:
根据目标社交网络平台中的用户信息构建无向连通图,其中,在所述无向连通图上,一个节点对应于一个用户,具有关联关系的用户之间通过一条边相连接;
获取由多个具有关联关系的目标用户构成的查询节点集,所述查询节点集为所述无向连通图中的节点集合的子集;
根据所述查询节点集和预设的社区搜索算法在所述无向连通图上进行迭代运算,以删除所述无向连通图上的节点,直至删除节点后得到满足第二预设条件的生成子图,该生成子图中的节点集合构成所述目标用户所属的社区;
将所述生成子图中的节点对应的用户作为所述多个目标用户的相似用户。
本申请提出的相似用户查找装置、方法及计算机可读存储介质,根据目标社交网络平台中的用户信息构建无向连通图,其中,在无向连通图上一个 节点对应于一个用户,具有关联关系的用户之间通过一条边相连接,获取由多个具有关联关系的目标用户构成的查询节点集,同时,这个查询节点集是五香连通图中的节点集合的子集,根据查询节点集合预设的社区搜索算法在无向连通图上进行迭代运算,以删除无向连通图上的节点,直至删除节点后得到满足第二预设条件的生成子图,该生成子图中的节点集合构成目标用户所属的社区,而在无向连通图上,属于一个社区的用户具有相似的属性,因此,该生成子图上的节点对应的用户可以作为上述目标用户的相似用户。本申请通过社交网络平台进行社区搜索,查找目标用户所述的社区,进而获取目标用户的相似用户,提高了查找相似用户的准确度。
附图说明
图1为本申请基于社区搜索的相似用户查找装置较佳实施例的示意图;
图2为本申请基于社区搜索的相似用户查找装置一实施例中用户查找程序的程序模块示意图;
图3为本申请基于社区搜索的相似用户查找方法第一实施例的流程图。
本申请目的的实现、功能特点及优点将结合实施例,参照附图做进一步说明。
具体实施方式
应当理解,此处所描述的具体实施例仅仅用以解释本申请,并不用于限定本申请。
本申请提供一种基于社区搜索的相似用户查找装置。参照图1所示,为本申请基于社区搜索的相似用户查找装置较佳实施例的示意图。
在本实施例中,基于社区搜索的相似用户查找装置1可以是PC(Personal Computer,个人电脑),也可以是智能手机、平板电脑、便携计算机等终端设备。该基于社区搜索的相似用户查找装置1至少包括存储器11、处理器12,网络接口13,以及通信总线14。
其中,存储器11至少包括一种类型的可读存储介质,所述可读存储介质包括闪存、硬盘、多媒体卡、卡型存储器(例如,SD或DX存储器等)、磁性存储器、磁盘、光盘等。存储器11在一些实施例中可以是基于社区搜索的相似用户查找装置1的内部存储单元,例如该基于社区搜索的相似用户查找装置1的硬盘。存储器11在另一些实施例中也可以是基于社区搜索的相似用 户查找装置1的外部存储设备,例如基于社区搜索的相似用户查找装置1上配备的插接式硬盘,智能存储卡(Smart Media Card,SMC),安全数字(Secure Digital,SD)卡,闪存卡(Flash Card)等。进一步地,存储器11还可以既包括基于社区搜索的相似用户查找装置1的内部存储单元也包括外部存储设备。存储器11不仅可以用于存储安装于基于社区搜索的相似用户查找装置1的应用软件及各类数据,例如用户查找程序01的代码等,还可以用于暂时地存储已经输出或者将要输出的数据。
处理器12在一些实施例中可以是一中央处理器(Central Processing Unit,CPU)、控制器、微控制器、微处理器或其他数据处理芯片,用于运行存储器11中存储的程序代码或处理数据,例如执行用户查找程序01等。
网络接口13可选的可以包括标准的有线接口、无线接口(如WI-FI接口),通常用于在该装置1与其他电子设备之间建立通信连接。
通信总线14用于实现这些组件之间的连接通信。
图1仅示出了具有组件11-14以及用户查找程序01的基于社区搜索的相似用户查找装置1,但是应理解的是,并不要求实施所有示出的组件,可以替代的实施更多或者更少的组件。
可选地,该装置1还可以包括用户接口,用户接口可以包括显示器(Display)、输入单元比如键盘(Keyboard),可选的用户接口还可以包括标准的有线接口、无线接口。可选地,在一些实施例中,显示器可以是LED显示器、液晶显示器、触控式液晶显示器以及OLED(Organic Light-Emitting Diode,有机发光二极管)触摸器等。其中,显示器也可以适当的称为显示屏或显示单元,用于显示在基于社区搜索的相似用户查找装置1中处理的信息以及用于显示可视化的用户界面。
在图1所示的装置1实施例中,存储器11中存储有用户查找程序01;处理器12执行存储器11中存储的用户查找程序01时实现如下步骤:
根据目标社交网络平台中的用户信息构建无向连通图,其中,在所述无向连通图上,一个节点对应于一个用户,具有关联关系的用户之间通过一条边相连接。
获取一个社交网络平台上预设数量用户的用户信息,用户信息中除了包含用户的个人资料之外,还包含用户的关注者信息和被关注者信息,这些信 息体现了用户之间的关联关系,其中,关注关系可以是用户在上述平台上的关注关系、好友关系、共同投资等。其中,关注关系可以单向关注关系或者双向关注关系。基于用户之间的关联关系构建一个复杂网络,在该复杂网络上,一个节点对应于一个用户,具有关联关系的用户之间通过一条边相连接。该复杂网络可以抽象表示成一个无向连通图G=V,E),其中,V是用户节点的集合,E是边的集合,体现用户之间的关联关系,例如边e=(v 1,v 2)代表用户v1和用户v2之间的关系。本实施例中,社交网络平台包括微博平台、Twitter平台或者金融论坛等。
获取由多个具有关联关系的目标用户构成的查询节点集,所述查询节点集为所述无向连通图中的节点集合的子集。
根据所述查询节点集和预设的社区搜索算法在所述无向连通图上进行迭代运算,以删除所述无向连通图上的节点,直至删除节点后得到满足第二预设条件的生成子图,该生成子图中的节点集合构成所述目标用户所属的社区。
将所述生成子图中的节点对应的用户作为所述多个目标用户的相似用户。
获取多个目标用户,这些目标用户构成的查询节点集,查询节点集是图G的用户节点集V的一个子集。接下来根据社区搜索算法在无向连通图G上进行迭代计算,以找到与这些目标用户关系密切或者具有相似偏好的用户群。这些用户群以及他们之间的关联关系形成无向连通图G的一个生成子图。
具体地,根据所述查询节点集和预设的社区搜索算法在所述无向连通图上进行迭代运算,以删除所述无向连通图上的节点,直至删除节点后得到满足第二预设条件的生成子图的步骤包括:
删除无向连通图中具有最小度的节点,以及与该节点相连的边;判断经节点删除操作后的无向连通图是否满足第二预设条件,其中第二预设条件为:所述查询节点集中的节点在节点删除操作后的图无向连通图中具有最小度,或者,经节点删除操作后,查询节点集中的节点在该无向连通图上不再连通;若是,则终止节点删除操作,将当前的无向连通图作为所述生成子图;若否,则基于当前的无向连通图继续执行所述删除无向连通图中具有最小度的节点,以及与该节点相连的边的步骤。
该社区搜索算法主要包括:将G0作为算法的初始输入图,其中,G0=G;每一步删除该步骤中输入的图上的一个节点以及与该节点关联的边,该节点 是当前的图中具有最小度的节点,其中,在无向连通图中,节点的度等于连接该节点的边的数量。即在第t步中,计算当前的图Gt-1上具有最小度的节点,删除该节点以及所有与该节点连接的边,得到图Gt,作为下一步的输入。如此循环往复的删除每一步骤的图中具有最小度的节点,直至执行了节点删除操作后得到的图满足以下条件之一:查询节点集中的节点在图中具有最小度,或者,删除有最小度的节点后,查询节点集中的点不再连通。这时,终止节点删除操作,获取计算终止后得到的生成子图,该生成子图是一个连通图,并且查询节点集是该生成子图中节点集的一个子集。
按照上述算法获取到的生成子图中的节点对应的用户即为要搜索的社区,这些用户具有与目标用户相同的特征,例如,具有相同的感兴趣产品或者具有相同的特征标签等。
进一步地,在其他的实施例中,为了避免计算得到的社区范围过大,可以进一步地增加预设条件:生成子图中的节点与查询节点集Q中的节点的最大距离不超过预设距离。在判定经节点删除操作后的无向连通图满足第二预设条件之后,终止节点删除操作之前,计算当前的无向连通图中的节点到所述查询节点集之间的距离,判断是否有节点到查询节点的距离大于预设距离;若是,则删除到查询节点的距离大于预设距离的节点,并将删除节点后的无向连通图作为所述生成子图;若否,则执行将终止节点删除操作,将当前的无向连通图作为所述生成子图的步骤。
关于节点之间的距离的计算可以采用如下方式中的任意一种:方式一,将两个节点之间的最短路径中的边的数量作为节点之间的距离,由于在一个复杂的无向连通图中,可以由多种路径实现两节点之间的连通,此处可以根据多条路径中的最短路径来计算节点之间的距离;方式二,根据两个节点之间的最短路径中的边的权重计算节点之间的距离,具体地,从目标社交网络平台上获取到的用户信息中包含有标签信息,根据两个用户拥有的相同标签的数量设置他们之间的权重,相同标签的数量越多,则权重越大,否则权重越小,将两个节点之间的最短路径中的边的权重之和作为两个节点之间的距离。其中,标签信息可以是用户在社交网络平台上可以为自己添加的标签,这些标签体现用户的感兴趣标签信息,或者,也可以将从用户在该社交网路平台上发布的内容提取的关键字作为用户的标签信息。此外,在计算一个节 点到查询节点集的距离时,先计算该节点到查询节点集中每一个节点的距离,然后将上述多个距离中的最小距离作为该节点到查询节点集的距离。
此外,可以理解的是,在上述删除节点的过程中,如果删除了某个节点之后,有节点在图上变成孤立的点,与其他节点之间没有任何连接关系时,要把这些孤立的节点删除。
通过上述计算获取到目标用户所在的社区,即由生成子图中的节点对应的用户构成的社区,这些用户是目标用户的相似用户,例如,具体在一个金融论坛中,可以认为在同一个社区中的用户有相同或者相似的投资偏好、属性、标签、感兴趣事件等。在得到相似用户后,根据它们之间相同或者相似的投资偏好、属性、标签、感兴趣事件等信息做精准推荐或定向投放。
以上实施例提出的相似用户查找装置,根据目标社交网络平台中的用户信息构建无向连通图,其中,在无向连通图上一个节点对应于一个用户,具有关联关系的用户之间通过一条边相连接,获取由多个具有关联关系的目标用户构成的查询节点集,同时,这个查询节点集是五香连通图中的节点集合的子集,根据查询节点集合预设的社区搜索算法在无向连通图上进行迭代运算,以删除无向连通图上的节点,直至删除节点后得到满足第二预设条件的生成子图,该生成子图中的节点集合构成目标用户所属的社区,而在无向连通图上,属于一个社区的用户具有相似的属性,因此,该生成子图上的节点对应的用户可以作为上述目标用户的相似用户。本申请通过社交网络平台进行社区搜索,查找目标用户所述的社区,进而获取目标用户的相似用户,提高了查找相似用户的准确度。
可选地,在其他的实施例中,用户查找程序还可以被分割为一个或者多个模块,一个或者多个模块被存储于存储器11中,并由一个或多个处理器(本实施例为处理器12)所执行以完成本申请,本申请所称的模块是指能够完成特定功能的一系列计算机程序指令段,用于描述用户查找程序在基于社区搜索的相似用户查找装置中的执行过程。
例如,参照图2所示,为本申请基于社区搜索的相似用户查找装置一实施例中的用户查找程序的程序模块示意图,该实施例中,用户查找程序可以被分割为信息生成模块10、信息获取模块20、社区搜索模块30、用户获取模块40,示例性地:
信息生成模块10用于:根据目标社交网络平台中的用户信息构建无向连通图,其中,在所述无向连通图上,一个节点对应于一个用户,具有关联关系的用户之间通过一条边相连接;
信息获取模块20用于:获取由多个具有关联关系的目标用户构成的查询节点集,所述查询节点集为所述无向连通图中的节点集合的子集;
社区搜索模块30用于:根据所述查询节点集和预设的社区搜索算法在所述无向连通图上进行迭代运算,以删除所述无向连通图上的节点,直至删除节点后得到满足第二预设条件的生成子图,该生成子图中的节点集合构成所述目标用户所属的社区;
用户获取模块40用于:将所述生成子图中的节点对应的用户作为所述多个目标用户的相似用户。
上述信息生成模块10、信息获取模块20、社区搜索模块30、用户获取模块40等程序模块被执行时所实现的功能或操作步骤与上述实施例大体相同,在此不再赘述。
此外,本申请还提供一种基于社区搜索的相似用户查找方法。参照图3所示,为本申请基于社区搜索的相似用户查找方法第一实施例的流程图。该方法可以由一个装置执行,该装置可以由软件和/或硬件实现。
在本实施例中,基于社区搜索的相似用户查找方法包括:
步骤S10,根据目标社交网络平台中的用户信息构建无向连通图,其中,在所述无向连通图上,一个节点对应于一个用户,具有关联关系的用户之间通过一条边相连接。
获取一个社交网络平台上预设数量用户的用户信息,用户信息中除了包含用户的个人资料之外,还包含用户的关注者信息和被关注者信息,这些信息体现了用户之间的关联关系,其中,关注关系可以是用户在上述平台上的关注关系、好友关系、共同投资等。其中,关注关系可以单向关注关系或者双向关注关系。基于用户之间的关联关系构建一个复杂网络,在该复杂网络上,一个节点对应于一个用户,具有关联关系的用户之间通过一条边相连接。该复杂网络可以抽象表示成一个无向连通图G=V,E),其中,V是用户节点的集合,E是边的集合,体现用户之间的关联关系,例如边e=(v 1,v 2)代表用户v1和用户v2之间的关系。本实施例中,社交网络平台包括微博平台、Twitter 平台或者金融论坛等。
步骤S20,获取由多个具有关联关系的目标用户构成的查询节点集,所述查询节点集为所述无向连通图中的节点集合的子集。
步骤S30,根据所述查询节点集和预设的社区搜索算法在所述无向连通图上进行迭代运算,以删除所述无向连通图上的节点,直至删除节点后得到满足第二预设条件的生成子图,该生成子图中的节点集合构成所述目标用户所属的社区。
步骤S40,将所述生成子图中的节点对应的用户作为所述多个目标用户的相似用户。
获取多个目标用户,这些目标用户构成的查询节点集,查询节点集是图G的用户节点集V的一个子集。接下来根据社区搜索算法在无向连通图G上进行迭代计算,以找到与这些目标用户关系密切或者具有相似偏好的用户群。这些用户群以及他们之间的关联关系形成无向连通图G的一个生成子图。
具体地,步骤S30包括:
删除无向连通图中具有最小度的节点,以及与该节点相连的边;判断经节点删除操作后的无向连通图是否满足第二预设条件,其中第二预设条件为:所述查询节点集中的节点在节点删除操作后的图无向连通图中具有最小度,或者,经节点删除操作后,查询节点集中的节点在该无向连通图上不再连通;若是,则终止节点删除操作,将当前的无向连通图作为所述生成子图;若否,则基于当前的无向连通图继续执行所述删除无向连通图中具有最小度的节点,以及与该节点相连的边的步骤。
该社区搜索算法主要包括:将G0作为算法的初始输入图,其中,G0=G;每一步删除该步骤中输入的图上的一个节点以及与该节点关联的边,该节点是当前的图中具有最小度的节点,其中,在无向连通图中,节点的度等于连接该节点的边的数量。即在第t步中,计算当前的图Gt-1上具有最小度的节点,删除该节点以及所有与该节点连接的边,得到图Gt,作为下一步的输入。如此循环往复的删除每一步骤的图中具有最小度的节点,直至执行了节点删除操作后得到的图满足以下条件之一:查询节点集中的节点在图中具有最小度,或者,删除有最小度的节点后,查询节点集中的点不再连通。这时,终止节点删除操作,获取计算终止后得到的生成子图,该生成子图是一个连通 图,并且查询节点集是该生成子图中节点集的一个子集。
按照上述算法获取到的生成子图中的节点对应的用户即为要搜索的社区,这些用户具有与目标用户相同的特征,例如,具有相同的感兴趣产品或者具有相同的特征标签等。
进一步地,在其他的实施例中,为了避免计算得到的社区范围过大,可以进一步地增加预设条件:生成子图中的节点与查询节点集Q中的节点的最大距离不超过预设距离。在判定经节点删除操作后的无向连通图满足第二预设条件之后,终止节点删除操作之前,计算当前的无向连通图中的节点到所述查询节点集之间的距离,判断是否有节点到查询节点的距离大于预设距离;若是,则删除到查询节点的距离大于预设距离的节点,并将删除节点后的无向连通图作为所述生成子图;若否,则执行将终止节点删除操作,将当前的无向连通图作为所述生成子图的步骤。
关于节点之间的距离的计算可以采用如下方式中的任意一种:方式一,将两个节点之间的最短路径中的边的数量作为节点之间的距离,由于在一个复杂的无向连通图中,可以由多种路径实现两节点之间的连通,此处可以根据多条路径中的最短路径来计算节点之间的距离;方式二,根据两个节点之间的最短路径中的边的权重计算节点之间的距离,具体地,从目标社交网络平台上获取到的用户信息中包含有标签信息,根据两个用户拥有的相同标签的数量设置他们之间的权重,相同标签的数量越多,则权重越大,否则权重越小,将两个节点之间的最短路径中的边的权重之和作为两个节点之间的距离。其中,标签信息可以是用户在社交网络平台上可以为自己添加的标签,这些标签体现用户的感兴趣标签信息,或者,也可以将从用户在该社交网路平台上发布的内容提取的关键字作为用户的标签信息。此外,在计算一个节点到查询节点集的距离时,先计算该节点到查询节点集中每一个节点的距离,然后将上述多个距离中的最小距离作为该节点到查询节点集的距离。
此外,可以理解的是,在上述删除节点的过程中,如果删除了某个节点之后,有节点在图上变成孤立的点,与其他节点之间没有任何连接关系时,要把这些孤立的节点删除。
通过上述计算获取到目标用户所在的社区,即由生成子图中的节点对应的用户构成的社区,这些用户是目标用户的相似用户,例如,具体在一个金 融论坛中,可以认为在同一个社区中的用户有相同或者相似的投资偏好、属性、标签、感兴趣事件等。在得到相似用户后,根据它们之间相同或者相似的投资偏好、属性、标签、感兴趣事件等信息做精准推荐或定向投放。
以上实施例提出的相似用户查找方法,根据目标社交网络平台中的用户信息构建无向连通图,其中,在无向连通图上一个节点对应于一个用户,具有关联关系的用户之间通过一条边相连接,获取由多个具有关联关系的目标用户构成的查询节点集,同时,这个查询节点集是五香连通图中的节点集合的子集,根据查询节点集合预设的社区搜索算法在无向连通图上进行迭代运算,以删除无向连通图上的节点,直至删除节点后得到满足第二预设条件的生成子图,该生成子图中的节点集合构成目标用户所属的社区,而在无向连通图上,属于一个社区的用户具有相似的属性,因此,该生成子图上的节点对应的用户可以作为上述目标用户的相似用户。本申请通过社交网络平台进行社区搜索,查找目标用户所述的社区,进而获取目标用户的相似用户,提高了查找相似用户的准确度。
此外,本申请实施例还提出一种计算机可读存储介质,所述计算机可读存储介质上存储有用户查找程序,所述用户查找程序可被一个或多个处理器执行,以实现如下操作:
根据目标社交网络平台中的用户信息构建无向连通图,其中,在所述无向连通图上,一个节点对应于一个用户,具有关联关系的用户之间通过一条边相连接;
获取由多个具有关联关系的目标用户构成的查询节点集,所述查询节点集为所述无向连通图中的节点集合的子集;
根据所述查询节点集和预设的社区搜索算法在所述无向连通图上进行迭代运算,以删除所述无向连通图上的节点,直至删除节点后得到满足第二预设条件的生成子图,该生成子图中的节点集合构成所述目标用户所属的社区;
将所述生成子图中的节点对应的用户作为所述多个目标用户的相似用户。
本申请计算机可读存储介质具体实施方式与上述基于社区搜索的相似用户查找装置和方法各实施例基本相同,在此不作累述。
需要说明的是,上述本申请实施例序号仅仅为了描述,不代表实施例的优劣。并且本文中的术语“包括”、“包含”或者其任何其他变体意在涵盖非 排他性的包含,从而使得包括一系列要素的过程、装置、物品或者方法不仅包括那些要素,而且还包括没有明确列出的其他要素,或者是还包括为这种过程、装置、物品或者方法所固有的要素。在没有更多限制的情况下,由语句“包括一个……”限定的要素,并不排除在包括该要素的过程、装置、物品或者方法中还存在另外的相同要素。
通过以上的实施方式的描述,本领域的技术人员可以清楚地了解到上述实施例方法可借助软件加必需的通用硬件平台的方式来实现,当然也可以通过硬件,但很多情况下前者是更佳的实施方式。基于这样的理解,本申请的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来,该计算机软件产品存储在如上所述的一个存储介质(如ROM/RAM、磁碟、光盘)中,包括若干指令用以使得一台终端设备(可以是手机,计算机,服务器,或者网络设备等)执行本申请各个实施例所述的方法。
以上仅为本申请的优选实施例,并非因此限制本申请的专利范围,凡是利用本申请说明书及附图内容所作的等效结构或等效流程变换,或直接或间接运用在其他相关的技术领域,均同理包括在本申请的专利保护范围内。

Claims (20)

  1. 一种基于社区搜索的相似用户查找装置,其特征在于,所述装置包括存储器和处理器,所述存储器上存储有可在所述处理器上运行的用户查找程序,所述用户查找程序被所述处理器执行时实现如下步骤:
    根据目标社交网络平台中的用户信息构建无向连通图,其中,在所述无向连通图上,一个节点对应于一个用户,具有关联关系的用户之间通过一条边相连接;
    获取由多个具有关联关系的目标用户构成的查询节点集,所述查询节点集为所述无向连通图中的节点集合的子集;
    根据所述查询节点集和预设的社区搜索算法在所述无向连通图上进行迭代运算,以删除所述无向连通图上的节点,直至删除节点后得到满足第二预设条件的生成子图,该生成子图中的节点集合构成所述目标用户所属的社区;
    将所述生成子图中的节点对应的用户作为所述多个目标用户的相似用户。
  2. 如权利要求1所述的基于社区搜索的相似用户查找装置,其特征在于,所述根据所述查询节点集和预设的社区搜索算法在所述无向连通图上进行迭代运算,以删除所述无向连通图上的节点,直至删除节点后得到满足第二预设条件的生成子图的步骤包括:
    删除无向连通图中具有最小度的节点,以及与该节点相连的边;
    判断经节点删除操作后的无向连通图是否满足第二预设条件,其中第二预设条件为:所述查询节点集中的节点在节点删除操作后的图无向连通图中具有最小度,或者,经节点删除操作后,查询节点集中的节点在该无向连通图上不再连通;
    若是,则终止节点删除操作,将当前的无向连通图作为所述生成子图;
    若否,则基于当前的无向连通图继续执行所述删除无向连通图中具有最小度的节点,以及与该节点相连的边的步骤。
  3. 如权利要求2所述的基于社区搜索的相似用户查找装置,其特征在于,所述用户查找程序还可被所述处理器执行,以在所述终止节点删除操作,将当前的无向连通图作为所述生成子图的步骤之前,还实现如下步骤:
    若经节点删除操作后的无向连通图满足第二预设条件,则计算当前的无向连通图中的节点到所述查询节点集之间的距离,判断是否有节点到查询节 点的距离大于预设距离;
    若是,则删除到查询节点的距离大于预设距离的节点,并将删除节点后的无向连通图作为所述生成子图;
    若否,则执行将终止节点删除操作,将当前的无向连通图作为所述生成子图的步骤。
  4. 如权利要求3所述的基于社区搜索的相似用户查找装置,其特征在于,所述计算当前的无向连通图中的节点到所述查询节点集之间的距离的步骤包括:
    根据两个节点之间的最短路径中的边的数量或者权重,计算所述无向连通图中的节点到所述查询节点集之间的距离。
  5. 如权利要求1所述的基于社区搜索的相似用户查找装置,其特征在于,所述根据目标社交网络平台中的用户信息构建无向连通图的步骤包括:
    获取目标社交网络平台中的用户信息,根据获取的用户信息确定用户之间的关联关系;
    根据获取到的用户信息和用户之间的关联关系构建无向连通图。
  6. 如权利要求2所述的基于社区搜索的相似用户查找装置,其特征在于,所述根据目标社交网络平台中的用户信息构建无向连通图的步骤包括:
    获取目标社交网络平台中的用户信息,根据获取的用户信息确定用户之间的关联关系;
    根据获取到的用户信息和用户之间的关联关系构建无向连通图。
  7. 如权利要求3所述的基于社区搜索的相似用户查找装置,其特征在于,所述根据目标社交网络平台中的用户信息构建无向连通图的步骤包括:
    获取目标社交网络平台中的用户信息,根据获取的用户信息确定用户之间的关联关系;
    根据获取到的用户信息和用户之间的关联关系构建无向连通图。
  8. 一种基于社区搜索的相似用户查找方法,其特征在于,所述方法包括:
    根据目标社交网络平台中的用户信息构建无向连通图,其中,在所述无向连通图上,一个节点对应于一个用户,具有关联关系的用户之间通过一条边相连接;
    获取由多个具有关联关系的目标用户构成的查询节点集,所述查询节点 集为所述无向连通图中的节点集合的子集;
    根据所述查询节点集和预设的社区搜索算法在所述无向连通图上进行迭代运算,以删除所述无向连通图上的节点,直至删除节点后得到满足第二预设条件的生成子图,该生成子图中的节点集合构成所述目标用户所属的社区;
    将所述生成子图中的节点对应的用户作为所述多个目标用户的相似用户。
  9. 如权利要求8所述的基于社区搜索的相似用户查找方法,其特征在于,所述根据所述查询节点集和预设的社区搜索算法在所述无向连通图上进行迭代运算,以删除所述无向连通图上的节点,直至删除节点后得到满足第二预设条件的生成子图的步骤包括:
    删除无向连通图中具有最小度的节点,以及与该节点相连的边;
    判断经节点删除操作后的无向连通图是否满足第二预设条件,其中第二预设条件为:所述查询节点集中的节点在节点删除操作后的图无向连通图中具有最小度,或者,经节点删除操作后,查询节点集中的节点在该无向连通图上不再连通;
    若是,则终止节点删除操作,将当前的无向连通图作为所述生成子图;
    若否,则基于当前的无向连通图继续执行所述删除无向连通图中具有最小度的节点,以及与该节点相连的边的步骤。
  10. 如权利要求9所述的基于社区搜索的相似用户查找方法,其特征在于,所述终止节点删除操作,将当前的无向连通图作为所述生成子图的步骤之前,所述方法还包括如下步骤:
    若经节点删除操作后的无向连通图满足第二预设条件,则计算当前的无向连通图中的节点到所述查询节点集之间的距离,判断是否有节点到查询节点的距离大于预设距离;
    若是,则删除到查询节点的距离大于预设距离的节点,并将删除节点后的无向连通图作为所述生成子图;
    若否,则执行将终止节点删除操作,将当前的无向连通图作为所述生成子图的步骤。
  11. 如权利要求10所述的基于社区搜索的相似用户查找方法,其特征在于,所述计算当前的无向连通图中的节点到所述查询节点集之间的距离的步骤包括:
    根据两个节点之间的最短路径中的边的数量或者权重,计算所述无向连通图中的节点到所述查询节点集之间的距离。
  12. 如权利要求8所述的基于社区搜索的相似用户查找方法,其特征在于,所述根据目标社交网络平台中的用户信息构建无向连通图的步骤包括:
    获取目标社交网络平台中的用户信息,根据获取的用户信息确定用户之间的关联关系;
    根据获取到的用户信息和用户之间的关联关系构建无向连通图。
  13. 如权利要求9所述的基于社区搜索的相似用户查找方法,其特征在于,所述根据目标社交网络平台中的用户信息构建无向连通图的步骤包括:
    获取目标社交网络平台中的用户信息,根据获取的用户信息确定用户之间的关联关系;
    根据获取到的用户信息和用户之间的关联关系构建无向连通图。
  14. 如权利要求10所述的基于社区搜索的相似用户查找方法,其特征在于,所述根据目标社交网络平台中的用户信息构建无向连通图的步骤包括:
    获取目标社交网络平台中的用户信息,根据获取的用户信息确定用户之间的关联关系;
    根据获取到的用户信息和用户之间的关联关系构建无向连通图。
  15. 一种计算机可读存储介质,其特征在于,所述计算机可读存储介质上存储有用户查找程序,所述用户查找程序可被一个或者多个处理器执行,以实现如下步骤:
    根据目标社交网络平台中的用户信息构建无向连通图,其中,在所述无向连通图上,一个节点对应于一个用户,具有关联关系的用户之间通过一条边相连接;
    获取由多个具有关联关系的目标用户构成的查询节点集,所述查询节点集为所述无向连通图中的节点集合的子集;
    根据所述查询节点集和预设的社区搜索算法在所述无向连通图上进行迭代运算,以删除所述无向连通图上的节点,直至删除节点后得到满足第二预设条件的生成子图,该生成子图中的节点集合构成所述目标用户所属的社区;
    将所述生成子图中的节点对应的用户作为所述多个目标用户的相似用户。
  16. 如权利要求15所述的计算机可读存储介质,其特征在于,所述根据 所述查询节点集和预设的社区搜索算法在所述无向连通图上进行迭代运算,以删除所述无向连通图上的节点,直至删除节点后得到满足第二预设条件的生成子图的步骤包括:
    删除无向连通图中具有最小度的节点,以及与该节点相连的边;
    判断经节点删除操作后的无向连通图是否满足第二预设条件,其中第二预设条件为:所述查询节点集中的节点在节点删除操作后的图无向连通图中具有最小度,或者,经节点删除操作后,查询节点集中的节点在该无向连通图上不再连通;
    若是,则终止节点删除操作,将当前的无向连通图作为所述生成子图;
    若否,则基于当前的无向连通图继续执行所述删除无向连通图中具有最小度的节点,以及与该节点相连的边的步骤。
  17. 如权利要求16所述的计算机可读存储介质,其特征在于,所述用户查找程序可被一个或者多个处理器执行,以在所述终止节点删除操作,将当前的无向连通图作为所述生成子图的步骤之前,还实现如下步骤:
    若经节点删除操作后的无向连通图满足第二预设条件,则计算当前的无向连通图中的节点到所述查询节点集之间的距离,判断是否有节点到查询节点的距离大于预设距离;
    若是,则删除到查询节点的距离大于预设距离的节点,并将删除节点后的无向连通图作为所述生成子图;
    若否,则执行将终止节点删除操作,将当前的无向连通图作为所述生成子图的步骤。
  18. 如权利要求17所述的计算机可读存储介质,其特征在于,所述计算当前的无向连通图中的节点到所述查询节点集之间的距离的步骤包括:
    根据两个节点之间的最短路径中的边的数量或者权重,计算所述无向连通图中的节点到所述查询节点集之间的距离。
  19. 如权利要求15所述的计算机可读存储介质,其特征在于,所述根据目标社交网络平台中的用户信息构建无向连通图的步骤包括:
    获取目标社交网络平台中的用户信息,根据获取的用户信息确定用户之间的关联关系;
    根据获取到的用户信息和用户之间的关联关系构建无向连通图。
  20. 如权利要求16所述的计算机可读存储介质,其特征在于,所述根据目标社交网络平台中的用户信息构建无向连通图的步骤包括:
    获取目标社交网络平台中的用户信息,根据获取的用户信息确定用户之间的关联关系;
    根据获取到的用户信息和用户之间的关联关系构建无向连通图。
PCT/CN2018/102112 2018-04-26 2018-08-24 相似用户查找装置、方法及计算机可读存储介质 WO2019205373A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201810382302.8 2018-04-26
CN201810382302.8A CN108804516B (zh) 2018-04-26 2018-04-26 相似用户查找装置、方法及计算机可读存储介质

Publications (2)

Publication Number Publication Date
WO2019205373A1 true WO2019205373A1 (zh) 2019-10-31
WO2019205373A9 WO2019205373A9 (zh) 2021-04-01

Family

ID=64093027

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/102112 WO2019205373A1 (zh) 2018-04-26 2018-08-24 相似用户查找装置、方法及计算机可读存储介质

Country Status (2)

Country Link
CN (1) CN108804516B (zh)
WO (1) WO2019205373A1 (zh)

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111191021A (zh) * 2018-11-14 2020-05-22 北京嘀嘀无限科技发展有限公司 职业预测方法、装置、设备及计算机可读存储介质
CN110334159A (zh) * 2019-05-29 2019-10-15 苏宁金融服务(上海)有限公司 基于关系图谱的信息查询方法和装置
CN110413848B (zh) * 2019-07-19 2022-04-15 上海赜睿信息科技有限公司 一种数据检索方法、电子设备和计算机可读存储介质
CN110855487B (zh) * 2019-11-13 2022-02-18 腾讯科技(深圳)有限公司 网络用户相似度管理方法、装置及存储介质
CN111353103B (zh) * 2020-03-03 2023-11-03 京东科技控股股份有限公司 用于确定用户社群信息的方法和装置
CN111506823A (zh) * 2020-03-11 2020-08-07 平安科技(深圳)有限公司 信息推荐方法、装置及计算机设备
CN111612038B (zh) * 2020-04-24 2024-04-26 平安直通咨询有限公司上海分公司 异常用户检测方法及装置、存储介质、电子设备
CN112084423B (zh) * 2020-08-31 2023-12-12 腾讯科技(深圳)有限公司 一种信息发布平台推荐方法及装置
CN112559903B (zh) * 2020-12-15 2024-04-05 广州市贺氏办公设备有限公司 一种社区成员搜索方法、系统、装置及介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070226248A1 (en) * 2006-03-21 2007-09-27 Timothy Paul Darr Social network aware pattern detection
US8606787B1 (en) * 2010-09-15 2013-12-10 Google Inc. Social network node clustering system and method
CN104102675A (zh) * 2013-04-15 2014-10-15 中国人民大学 基于用户关系的博主兴趣社区发现方法
CN104731962A (zh) * 2015-04-03 2015-06-24 重庆邮电大学 一种社交网络中基于相似社团的好友推荐方法及系统
CN105721279A (zh) * 2016-01-15 2016-06-29 中国联合网络通信有限公司广东省分公司 一种电信网络用户的交往圈挖掘方法及系统

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102194149B (zh) * 2010-03-01 2012-12-05 中国人民解放军国防科学技术大学 社区发现方法
US9652875B2 (en) * 2012-10-29 2017-05-16 Yahoo! Inc. Systems and methods for generating a dense graph
CN104462260B (zh) * 2014-11-21 2018-07-10 深圳大学 一种基于k-核的社交网络中的社区搜索方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070226248A1 (en) * 2006-03-21 2007-09-27 Timothy Paul Darr Social network aware pattern detection
US8606787B1 (en) * 2010-09-15 2013-12-10 Google Inc. Social network node clustering system and method
CN104102675A (zh) * 2013-04-15 2014-10-15 中国人民大学 基于用户关系的博主兴趣社区发现方法
CN104731962A (zh) * 2015-04-03 2015-06-24 重庆邮电大学 一种社交网络中基于相似社团的好友推荐方法及系统
CN105721279A (zh) * 2016-01-15 2016-06-29 中国联合网络通信有限公司广东省分公司 一种电信网络用户的交往圈挖掘方法及系统

Also Published As

Publication number Publication date
CN108804516A (zh) 2018-11-13
WO2019205373A9 (zh) 2021-04-01
CN108804516B (zh) 2021-03-02

Similar Documents

Publication Publication Date Title
WO2019205373A1 (zh) 相似用户查找装置、方法及计算机可读存储介质
WO2019041521A1 (zh) 用户关键词提取装置、方法及计算机可读存储介质
JP5457434B2 (ja) モバイル装置により記録された画像のコンテンツに基づくデータアクセス
US8583659B1 (en) Labeling samples in a similarity graph
US11727053B2 (en) Entity recognition from an image
WO2019242144A1 (zh) 电子装置、偏好倾向预测方法和计算机可读存储介质
WO2018077181A1 (zh) 图中心性计算方法、装置及存储介质
WO2019085474A1 (zh) 计算引擎实现方法、电子装置及存储介质
CN110738577B (zh) 社区发现方法、装置、计算机设备和存储介质
US9734207B2 (en) Entity resolution techniques and systems
CN111125269B (zh) 一种数据管理方法、血缘关系显示方法和相关装置
AU2017268599B2 (en) Method, device, server and storage medium of searching a group based on social network
US9754015B2 (en) Feature rich view of an entity subgraph
WO2020047921A1 (zh) 基于层次三元组损失函数的深度度量学习方法及其装置
Cao et al. KORS: Keyword-aware optimal route search system
WO2019019385A1 (zh) 跨平台数据匹配方法、装置、计算机设备和存储介质
JP2017004252A (ja) 画像情報処理システム
CN112559717A (zh) 搜索匹配方法、装置、电子设备以及存储介质
WO2015024522A1 (zh) 搜索方法、系统、搜索引擎和客户端
WO2018227773A1 (zh) 地点推荐方法、装置、计算机设备和存储介质
WO2021027149A1 (zh) 基于画像相似性的信息检索推荐方法、装置及存储介质
TWI639093B (zh) Object set and processing method and device thereof
US10204139B2 (en) Systems and methods for processing geographic data
JP2017211846A (ja) 検索データ管理装置、検索データ管理方法、および検索データ管理プログラム
JP2014142803A (ja) 領域検索方法および領域検索装置

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: 18916250

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 18916250

Country of ref document: EP

Kind code of ref document: A1