Community determination method and apparatus, program product, and electronic device

Abstract

The invention discloses a community determination method and device, a program product and electronic equipment, and relates to the field of big data or other related fields, and the community determination method comprises the steps: determining an initial node in a financial network based on the business purpose of a financial business, constructing an initial community and a shell node set based on the initial node, and determining a shell node set based on the initial community; and based on the initial community and the shell node set, determining a target community in the financial network by adopting a preset second-order neighbor overlap ratio algorithm. According to the invention, the technical problem of low accuracy of community determination in a network in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the field of big data, and in particular to a community determination method and apparatus, a program product, and an electronic device thereof. Background Art

[0002] With the rapid development of internet technology, data has become increasingly massive and complex, and the resulting networks have become increasingly complex, making analysis of complex networks more difficult. However, in practical research, much of this data is useless. Often, studying a specific area can yield the desired results. For example, in a recommendation system, accurate recommendations can be made based on a person's interest groups, without having to study the entire interest network. Therefore, network analysis is currently typically performed using local community discovery models.

[0003] Figure 1 It is a schematic diagram of the local community discovery model in related technologies, such as Figure 1 As shown in Figure 1, the nodes in the network are divided into three parts: local communities D, shell node sets N, and unknown communities U. Local communities D are the set of nodes to be mined. Generally, local community discovery involves mining local communities D based on information in regions D and N, given some nodes in D. Node and edge information in the unknown community U is not required in most local community discovery algorithms.

[0004] In the related art, the following methods are generally used to determine local communities: (1) a local community determination method based on local modularity. However, the local community determination effect of this method is greatly affected by the selection of seed nodes; (2) a local community determination method based on the common neighbor similarity index. However, this common neighbor similarity index is only applicable to weighted networks and has a poor effect on local community determination in unweighted networks; (3) a local community determination method based on fuzzy similarity relationships. However, this method is applicable to networks with graph structures (i.e., highly interconnected node groups) and has a poor effect on local community determination in sparse networks.

[0005] To the above problems, no effective solution has been proposed yet. Summary of the Invention

[0006] Embodiments of the present invention provide a community determination method and apparatus, a program product, and an electronic device thereof, to at least solve the technical problem of low accuracy in community determination in a network in related technologies.

[0007] According to an aspect of the embodiments of the present application, a community determining method is provided, comprising: determining an initial node in a financial network based on a business purpose of a financial business, wherein the financial network is composed of a plurality of nodes, each of which indicates a user identity; constructing an initial community and a shell node set based on the initial node, wherein the initial community is a node set containing the initial node, and the shell node set contains nodes connected to the nodes in the initial community but not included in the initial community; and determining a target community in the financial network based on the initial community and the shell node set by using a preset second-order neighbor overlap degree algorithm, wherein each node in the target community indicates a user identity required for processing the financial business.

[0008] Further, the step of determining the initial node in the financial network based on the business purpose of the financial business comprises: determining at least one target user based on the business purpose of the financial business; and determining a node corresponding to a target user identity of the target user in the financial network as the initial node.

[0009] Further, the step of constructing the initial community and the shell node set based on the initial node comprises: constructing the initial community based on the initial node; querying neighbor nodes adjacent to the initial node in the financial network; and constructing the shell node set based on all the neighbor nodes.

[0010] Further, the step of determining the target community in the financial network based on the initial community and the shell node set by using the preset second-order neighbor overlap degree algorithm comprises: traversing the initial community, and selecting a node in the initial community as a community node; determining a second-order neighbor overlap degree between the community node and any neighbor node of the community node; in a case where the second-order neighbor overlap degree is greater than or equal to a preset overlap degree threshold and the neighbor node belongs to the shell node set, adding the neighbor node to the initial community and adding all adjacent nodes of the neighbor node to the shell node set; continuing to determine the second-order neighbor overlap degree between the selected community node and any neighbor node of the community node until the shell node set is empty or the second-order neighbor overlap degree between any node in the shell node set and the community node is all less than the preset overlap degree threshold, to obtain the target community.

[0011] Further, the step of determining the second-order neighbor overlap degree between the community node and any neighbor node of the community node comprises: determining a second-order neighbor node set of the community node and a second-order neighbor node set of the neighbor node; and determining the second-order neighbor overlap degree between the community node and the neighbor node based on the second-order neighbor node set of the community node and the second-order neighbor node set of the neighbor node.

[0012] Further, the step of determining the second-order neighbor node set of the community node and the second-order neighbor node set of the neighbor node comprises: determining a first node set composed of all neighbor nodes of the node, wherein the node is the community node or the neighbor node; determining the first-order neighbor node set based on the node and the first node set; determining a second node set composed of neighbor nodes of each first node in the first node set; and determining the second-order neighbor node set based on the first-order neighbor node set and all second node sets.

[0013] Further, the step of determining the second-order neighbor overlap degree between the community node and the neighbor node based on the second-order neighbor node set of the community node and the second-order neighbor node set of the neighbor node comprises: determining an intersection set between the second-order neighbor node set of the community node and the second-order neighbor node set of the neighbor node; determining a first node number of the second-order neighbor node set of the community node and a second node number of the second-order neighbor node set of the neighbor node; and determining the second-order neighbor overlap degree based on an intersection node number of the intersection set, the first node number and the second node number.

[0014] According to another aspect of the embodiment of the present application, a community determining apparatus is further provided, comprising: a first determining unit configured to determine an initial node in a financial network based on a business purpose of a financial business, wherein the financial network is composed of a plurality of nodes, and each node indicates a user identifier; a constructing unit configured to construct an initial community and a shell node set based on the initial node, wherein the initial community is a node set containing the initial node, and the shell node set contains nodes connected with nodes in the initial community but not included in the initial community; and a second determining unit configured to determine a target community in the financial network based on the initial community and the shell node set by using a preset second-order neighbor overlap degree algorithm, wherein each node in the target community indicates a user identifier required for processing the financial business.

[0015] Further, the first determining unit comprises: a first determining module configured to determine at least one target user based on the business purpose of the financial business; and a second determining module configured to determine a node corresponding to a target user identifier of the target user in the financial network as the initial node.

[0016] Further, the constructing unit comprises: a first constructing module configured to construct the initial community based on the initial node; a first querying module configured to query neighbor nodes adjacent to the initial node in the financial network; and a second constructing module configured to construct the shell node set based on all neighbor nodes.

[0017] Furthermore, the second determination unit includes: a first traversal module, configured to traverse the initial community and select nodes in the initial community as community nodes; a third determination module, configured to determine the second-order neighbor overlap between the community node and any neighbor node of the community node; a first adding module, configured to add the neighbor node to the initial community and add all adjacent nodes of the neighbor node to the shell node set when the second-order neighbor overlap is greater than or equal to a preset overlap threshold and the neighbor node belongs to the shell node set; and a fourth determination module, configured to continue determining the second-order neighbor overlap between the selected community node and any neighbor node of the community node until the shell node set is empty or the second-order neighbor overlap between any node in the shell node set and the community node is less than the preset overlap threshold, thereby obtaining a target community.

[0018] Furthermore, the third determination module includes: a first determination submodule, used to determine the second-order neighbor node set of the community node and determine the second-order neighbor node set of the neighbor node; a second determination submodule, used to determine the second-order neighbor overlap between the community node and the neighbor node based on the second-order neighbor node set of the community node and the second-order neighbor node set of the neighbor node.

[0019] Furthermore, the first determination submodule includes: a third determination submodule, used to determine a first node set consisting of all adjacent nodes of a node, wherein the node is a community node or a neighbor node; a fourth determination submodule, used to determine a first-order neighbor node set based on the node and the first node set; a fifth determination submodule, used to determine a second node set consisting of adjacent nodes of each first node in the first node set; and a sixth determination submodule, used to determine a second-order neighbor node set based on the first-order neighbor node set and all second node sets.

[0020] Furthermore, the second determination submodule includes: a seventh determination submodule, used to determine the intersection set between the second-order neighbor node set of the community node and the second-order neighbor node set of the neighbor node; an eighth determination submodule, used to determine the first node number of the second-order neighbor node set of the community node, and determine the second node number of the second-order neighbor node set of the neighbor node; a ninth determination submodule, used to determine the second-order neighbor overlap based on the intersection node number, the first node number, and the second node number of the intersection set.

[0021] According to another aspect of an embodiment of the present invention, a computer program product is provided, including a non-volatile computer-readable storage medium, wherein the non-volatile computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, any one of the above-mentioned community determination methods is implemented.

[0022] According to another aspect of the embodiments of the present application, an electronic device is provided, which includes one or more processors and a memory storing one or more programs, wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to implement any of the community determination methods described above.

[0023] In the present application, based on the business purpose of financial services, an initial node is determined in a financial network, based on the initial node, an initial community and a set of shell nodes are constructed, based on the initial community and the set of shell nodes, a target community is determined in the financial network by using a preset second-order neighbor overlap degree algorithm, thereby solving the technical problem of low accuracy of community determination in the network in the related art.

[0024] In the present application, by constructing an initial community and a set of shell nodes according to the determined initial node, the target community can be accurately divided in the financial network by using the preset second-order neighbor overlap degree algorithm, thus the user identifier related to the financial service can be effectively identified, the nodes of the target community are ensured to be highly matched with the business demand, the interference of invalid users is avoided, the execution quality and safety of the financial service are improved, the purpose of accurately positioning the target community in the financial network is achieved, and the technical effect of improving the business efficiency is achieved. BRIEF DESCRIPTION OF DRAWINGS

[0025] The accompanying drawings, which are included to provide a further understanding of the present application and are incorporated in and constitute a part of this application, illustrate embodiments of the present application and together with the description serve to explain the present application. In the drawings:

[0026] Figure 1 is a schematic diagram of a local community discovery model according to the related art;

[0027] Figure 2 shows a hardware structure block diagram of a computer terminal (or mobile device) for implementing the community determination method;

[0028] Figure 3 is a flowchart of the community determination method according to Embodiment 1 of the present application;

[0029] Figure 4 is a schematic diagram of a relatively sparse example network according to Embodiment 1 of the present application;

[0030] Figure 5 is a schematic diagram of an example network with second-order neighbor overlap degree according to Embodiment 1 of the present application;

[0031] Figure 6 is a schematic diagram of the SNC algorithm flow according to Embodiment 1 of the present application;

[0032] Figure 7is a schematic diagram of experimental results of different algorithms according to embodiment 1 of the present application on multiple real networks;

[0033] Figure 8 is a schematic diagram of experimental results of setting different q values of SNC algorithm according to embodiment 1 of the present application on multiple real networks;

[0034] Figure 9 is a schematic diagram of an optional community determining device according to an embodiment of the present application;

[0035] Figure 10 is a structural block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0036] In order to make the personnel in the technical field better understand the present application scheme, the technical scheme in the embodiments of the present application will be described clearly and completely below in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by the person skilled in the art without creative labor should belong to the scope of protection of the present application.

[0037] It should be noted that the terms "first", "second", and the like in the present application are used to distinguish similar objects, and do not necessarily indicate a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not necessarily limit to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0038] In order to facilitate the person skilled in the art to understand the present application, the following explains some terms or names involved in each embodiment of the present application:

[0039] Community concept:

[0040] In addition to the scale-free characteristics, complex networks also have a structure reflecting the tightness of the connection between some nodes in the network, which is called "community structure". The nodes in the network are often divided into some node groups, and the nodes in each node group usually have more similar structural characteristics, and the nodes in the same node group have a greater probability of generating edges. Such node group is called community.

[0041] In an undirected and unweighted network G(V, E) (V represents nodes and E represents edges), nodes in the network can be generally divided into k (k≥1) groups, each group being a community. Nodes in the same community are densely connected, and edges across communities are few. Community labels are generally represented by C, as shown in formula (1), and a network has k communities.

[0042] C={C1, C2, …, C k} (1);

[0043] Local community discovery concept:

[0044] Local community discovery is to mine a local community D in which a node (or multiple nodes) is located through the node and part of the network structure information. If a local community D is mined through an algorithm and a given node is in a real community C i The more similar, the better the effect of the local community discovery algorithm.

[0045] It should be noted that the information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) collected and related by the present application are all information and data authorized by the user or authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of related data comply with relevant laws, regulations and standards in relevant regions, necessary security measures are taken, do not violate public order and good customs, and provide corresponding operation portal for user selection authorization or refusal. For example, the system and the related user or institution are provided with an interface, and before obtaining the related information, the interface needs to send an acquisition request to the aforementioned user or institution, and after receiving the consent information fed back by the aforementioned user or institution, the related information is acquired; if the user chooses to refuse, the expert decision-making process is entered.

[0046] In the present application, the mining of relatively important nodes in a complex network has important research significance and a wide range of applications, such as finding remaining risk users in a financial network through known risk users. However, with the development of the times, the scale of the network is constantly expanding, and due to the excessively large scale of part of the network, the time spent on mining relatively important nodes may be large, and even the accuracy of the mining may be affected. However, the network is composed of a number of small communities, nodes in the same community not only have closer relationships between them but also have more similar functions, and the relatively important nodes to be mined will have a high probability of belonging to the same community as the known important nodes.

[0047] Therefore, in order to better mine relatively important nodes and narrow the research scope, the present invention proposes an algorithm for local community discovery that uses known important nodes as specific nodes, namely, a local community discovery algorithm based on the coincidence of known important nodes and second-order neighbors (Second-order Neighbor Coincidence, abbreviated as SNC).

[0048] The present invention will be described in detail below with reference to various embodiments.

[0049] Example 1

[0050] According to an embodiment of the present application, an embodiment of a community determination method is also provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0051] The method embodiment provided in the first embodiment of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Figure 2 FIG. 1 shows a hardware structure block diagram of a computer terminal (or mobile device) for implementing a community determination method. Figure 2 As shown, the computer terminal 20 (or mobile device) may include one or more ( Figure 2 202a, 202b, ..., 202n are used to illustrate) a processor 202 (the processor 202 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 204 for storing data, and a transmission device 206 for communication functions. In addition, it may also include: a display, a keyboard, a cursor control device, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the BUS bus), a network interface, a power supply and / or a camera, wherein the network interface may be connected to a wired and / or wireless network. It will be understood by those skilled in the art that Figure 2 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 2 More or fewer components than shown, or with Figure 2 Different configurations shown.

[0052] It should be noted that the one or more processors 202 and / or other data processing circuitry described above can be generally referred to herein as "data processing circuitry". The data processing circuitry can be embodied in whole or in part as software, hardware, firmware, or any combination thereof. In addition, the data processing circuitry can be a single standalone processing module, or incorporated in whole or in part within any of the other elements of the computer terminal 20 (or mobile device). As referred to in the embodiments of the present application, the data processing circuitry functions as a processor to control, for example, the selection of the variable resistance terminal path connected to the interface.

[0053] The memory 204 can be used to store software programs of application software and modules, such as program instructions / data storage means corresponding to the community determining method in the embodiments of the present application. The processor 202 executes various functional applications and data processing by running the software programs and modules stored in the memory 204, i.e. implements the community determining method described above. The memory 204 can include a high-speed random access memory, and can further include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some examples, the memory 204 can further include a memory remotely arranged with respect to the processor 202, which can be connected to the computer terminal 20 through a network. Examples of the network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0054] The transmission device 206 is used to receive or send data via a network. Specific examples of the network can include a wireless network provided by a communication provider of the computer terminal 20. In one example, the transmission device 206 includes a network adapter (NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 206 can be a radio frequency (RF) module, which is used to communicate with the Internet in a wireless manner.

[0055] The display can be, for example, a touch screen type liquid crystal display (LCD), which can enable a user to interact with the user interface of the computer terminal 20 (or mobile device).

[0056] Under the above operating environment, the present application provides a community determining method as shown in Figure 3 Figure 3 is a flowchart of the community determining method according to the embodiment 1 of the present application, which includes the following steps: Figure 3

[0057] ​​Step S301, based on the business purpose of the financial service, determining an initial node in the financial network, wherein the financial network is composed of a plurality of nodes, and each node indicates a user identifier.

[0058] In the embodiments of the present application, in the financial service scenario, the financial network is a complex structure composed of nodes and connection edges, and is a network constructed according to transaction data, consumption data and other data between users. The node in the financial network refers to a user or an account in the financial service, and each node is represented by a user identifier, which can be a user ID (identifier), an account name or other unique identifier.

[0059] In the embodiments of the present application, one or more representative nodes can be selected from the financial network according to the specific purpose of the financial service (such as risk control, personalized service, accurate recommendation, etc.) as the initial node. For example, in the risk control scenario, a known risk account can be used as the initial node, and the potential risk community can be identified based on the initial node.

[0060] Step S302, based on the initial node, constructing an initial community and a set of shell nodes, wherein the initial community is a set of nodes containing the initial node, and the set of shell nodes includes nodes connected to the nodes in the initial community but not included in the initial community.

[0061] In the embodiments of the present application, according to the known important node (i.e. the initial node) r, a local community node set D (i.e. the initial community, which is a set of nodes containing the initial node) and a shell node set N (including nodes connected to the nodes in the initial community but not included in the initial community) are constructed, wherein D={r}, N=Γ(r) (representing a set of nodes connected to the initial node in the initial community).

[0062] In the embodiments of the present application, constructing the initial community means extracting a set of closely connected nodes from the financial network, which have direct connection relationship with the initial node and share similar attributes or behavior patterns. The construction of the initial community essentially reduces the scope of network analysis to the part directly related to the business purpose, thereby improving the efficiency and accuracy of subsequent analysis. At the same time, the set of shell nodes includes nodes that have direct contact with the nodes in the initial community but have not been included in the community. These nodes have potential community attribution, and through further algorithm processing, it can be determined whether these nodes are finally included in the target community, which is a dynamic expansion process in the community discovery algorithm, ensuring the accuracy and integrity of community division.

[0063] In step S303, based on the initial community and the set of shell nodes, a preset second-order neighbor overlap degree algorithm is used to determine a target community in the financial network, wherein each node in the target community indicates a user identifier required for processing the financial service.

[0064] In the embodiments of the present application, the preset second-order neighbor overlap degree algorithm can be used to identify deeper connections in the financial network, that is, by analyzing the overlap degree between the neighbors of the neighbors (second-order neighbors) of the nodes to determine which shell nodes should be added to the target community. The second-order neighbor overlap degree algorithm measures the degree of overlap of the second-order neighbor node sets of nodes x and y, not only considering the direct connection relationship, but also considering the indirect contact strength, so that the user identifiers related to the business purpose can be more accurately identified. The determination of the target community means that all user identifier sets required for processing the specific financial service have been screened out, and the members in this community have high relevance and homogeneity, which is crucial for targeted operations of the business. For example, in risk control, the identification of the target community can help the financial institution to more accurately locate the potential risk users, so as to take more effective risk management and prevention measures.

[0065] In summary, by constructing the initial community and its shell node set according to the determined initial node, the preset second-order neighbor overlap degree algorithm can be used to accurately divide the target community in the financial network. In this way, the user identifiers related to the financial service can be effectively identified, the nodes of the target community are highly matched with the business requirements, the interference of invalid users is avoided, the execution quality and safety of the financial service are improved, and the purpose of accurately positioning the target community in the financial network is achieved, thereby improving the technical effect of business efficiency.

[0066] In order to improve the accuracy of determining the initial node, in the community determination method provided in Embodiment 1 of the present application, at least one target user is determined based on the business purpose of the financial service; and a node corresponding to a target user identifier of the target user in the financial network is determined as the initial node.

[0067] In the embodiments of the present application, in the practice of the financial service, the business purpose can be diverse, including but not limited to risk assessment, market segmentation, customer relationship management, etc. Taking risk assessment as an example, the business purpose can be to identify potential fraudulent behavior or credit risk. In this context, the target user refers to those users who have exhibited certain risk characteristics or are considered to be associated with risk behavior. For example, the target user can be determined by historical data analysis, behavior pattern recognition, and determined abnormal transaction records. The information of these users is the "seed" for the start of the community discovery algorithm, and through these target users, a more extensive potential risk group can be identified in the financial network.

[0068] In the embodiments of the present application, after the target users are determined, the nodes corresponding to these users in the financial network can be found and marked as initial nodes. The nodes in the financial network represent users or accounts, and each node has a user identifier (such as an account ID, etc.) associated therewith for uniquely identifying an individual in the network. Mapping the target user identifiers to the nodes of the financial network is essentially to convert the business-level focus into algorithmically processable data entities. In this way, the algorithm can start from these nodes with clear business orientation, explore the community structure associated with these nodes, and thus more effectively achieve the business purpose.

[0069] Exemplarily, user data related to the financial business target can be collected, and a target user list can be determined through data analysis. Subsequently, the information of the target users is imported into the financial network model, the network nodes corresponding to the target user identifiers are found, and these nodes are marked as initial nodes.

[0070] In the embodiments, according to the real-time business requirements and the behaviors of the target users, the focus of the financial network analysis is flexibly adjusted, so that the algorithm can not only quickly respond to business changes, but also efficiently focus on key user groups in a large network, which is beneficial to improving the risk management capability of the financial institution, optimizing the customer service experience, and enhancing the market competitiveness.

[0071] In order to improve the accuracy of constructing the initial community and the shell node set, in the community determination method provided in Embodiment 1 of the present application, based on the initial nodes, an initial community is constructed; neighbor nodes adjacent to the initial nodes in the financial network are queried; and based on all the neighbor nodes, a shell node set is constructed.

[0072] In the embodiments of the present application, the construction of the initial community is based on the determined initial nodes. The initial nodes are the nodes corresponding to the key users with specific business purposes in the financial network. The initial nodes can be added to the initial community to query a node set having similar characteristics with the initial nodes according to the initial nodes.

[0073] In the embodiments of the present application, all other nodes (i.e., neighbor nodes) directly connected to the nodes in the initial community in the financial network can be identified. By querying the neighbor nodes, user information having direct business dealings or associations with the initial nodes can be obtained, and these information is crucial for understanding the interaction patterns within the community. The query of the neighbor nodes includes not only the nodes directly connected to the initial nodes, but also the attributes of these nodes and the connection strength therebetween.

[0074] In the embodiment of the present application, all the neighbor nodes adjacent to the initial node can constitute a shell node set, i.e., the shell node set includes all the nodes having direct connection with the nodes in the initial community but not yet being classified into the initial community. These nodes are at the edge of the community, and the connection with the core nodes (i.e., the nodes in the initial community) constitutes the boundary condition of the community. By defining the shell node set, the expansion range of the community can be clearly defined, and an explicit search boundary is provided for the subsequent community discovery algorithm. In this way, not only the efficient operation of the algorithm is facilitated, but also the accuracy and integrity of the finally determined target community are guaranteed.

[0075] In the embodiment, by constructing the initial community, the starting point of the business analysis is determined, then the neighbor nodes are queried to further refine the boundary and internal structure of the community, and finally the shell node set is defined to provide an explicit guidance for the expansion of the community. In this way, not only the financial institution can more accurately identify and locate the specific user group, but also the business process can be optimized on this basis to improve the efficiency and quality of business processing, which has important practical value for financial risk control, customer relationship management and market strategy formulation.

[0076] In order to improve the accuracy of determining the target community, in the community determination method provided in Embodiment 1 of the present application, the initial community is traversed, and the selected node in the initial community is taken as a community node; the second-order neighbor coincidence degree between the community node and any neighbor node of the community node is determined; in the case that the second-order neighbor coincidence degree is greater than or equal to a preset coincidence degree threshold and the neighbor node belongs to the shell node set, the neighbor node is added to the initial community, and all the adjacent nodes of the neighbor node are added to the shell node set; the second-order neighbor coincidence degree between the selected community node and any neighbor node of the community node is continuously determined until the shell node set is empty or the second-order neighbor coincidence degree between any node in the shell node set and the community node is all less than the preset coincidence degree threshold, and the target community is obtained.

[0077] In the embodiment of the present application, since the degree distribution of a large part of the network in the abstract network of the complex system obeys the power-law distribution, to improve the division effect of the local community discovery on the small-degree nodes (i.e. the nodes with small number of connected edges), the local community discovery algorithm cannot be simply designed based on the graph structure (the structure of three nodes interconnected with each other). Therefore, the embodiment proposes a local community discovery algorithm based on the known important nodes and the second-order neighbor overlap degree. In the algorithm, the second-order neighbor overlap degree is used to replace the graph structure to measure the similarity between nodes, and the known important nodes are used as the specific nodes to mine the nodes whose second-order neighbor overlap degrees with respect to the known important nodes exceed the limit level, and the nodes are divided into the same local community. Then, the nodes whose second-order neighbor overlap degrees with respect to the nodes in the local community exceed the limit level are continuously mined and divided into the same community. The above steps are repeated until there is no such node.

[0078] In the embodiment of the present application, the local community discovery algorithm based on the known important nodes and the second-order neighbor overlap degree (SNC) adopts the method of the maximum connected subgraph to find the related community satisfying the condition. The algorithm needs to specify two parameters r and q, wherein the parameter r is the known important node (i.e. the initial node), and the parameter q is the second-order neighbor overlap degree limit coefficient (i.e. the preset overlap degree threshold, which can be regarded as a kind of representation of the node similarity). The SNC algorithm obtains the node set (i.e. the target community) belonging to the same community with the known important node r according to the known important node r and the second-order neighbor overlap degree coefficient q.

[0079] In the embodiment of the present application, the node x whose second-order neighbor overlap degree C(x, y) value is greater than or equal to q and the node y is located in the shell node set N is searched in the local community node set D (i.e. the initial community). If there is such an overlap degree relationship between x and y, the node y is added to the local community D. Then, the neighbor node of the node y newly added to the local community node set D is added to the shell node set N. The above steps are repeated until the shell node set N is empty, or the node y in the shell node set does not satisfy C(x, y) ≥ q.

[0080] Specifically, the initial community can be traversed first, i.e., the nodes in the community are checked one by one to determine the role of these nodes as community nodes in the subsequent algorithm. In this way, it can be ensured that all nodes in the initial community will participate in the calculation of the second-order neighbor overlap degree in the subsequent algorithm, so as to accurately identify the possibility of community expansion. Then, the degree of overlap between the community nodes and the second-order neighbor set of their neighbor nodes (i.e., the neighbor nodes of the neighbor nodes) (i.e., the second-order neighbor overlap degree) can be compared. This indicator can reflect the indirect association strength between nodes, and compared with the direct neighbor relationship, it can more comprehensively evaluate the closeness between nodes, especially when identifying community structures hidden at deeper levels. By calculating the second-order neighbor overlap degree, it can be identified which neighbor nodes have a higher indirect connection with the community nodes. If the calculated second-order neighbor overlap degree exceeds the preset overlap threshold, and the neighbor node belongs to the shell node set, the neighbor node will be absorbed into the initial community, and the direct neighbors of the neighbor node (i.e., all adjacent nodes) will be added to the shell node set for subsequent community expansion analysis. Then, by continuously evaluating the second-order neighbor overlap degree between the community nodes and the nodes in the shell node set, the community boundary is dynamically expanded until the stopping condition is met (i.e., the shell node set is empty, or the second-order neighbor overlap degree of all remaining shell nodes (i.e., the nodes in the shell node set) with the community nodes is lower than the preset overlap threshold). In this way, the algorithm can gradually and accurately define the scope of the target community, ensuring that the final target community is formed based on sufficient indirect connections between nodes, while avoiding the problem of unclear community structure caused by excessive expansion.

[0081] Exemplarily, the SNC algorithm can be represented as follows:

[0082]

[0083]

[0084] In this embodiment, by iteratively calculating the second-order neighbor overlap degree and dynamically adjusting the community boundary, a user group highly related to the business target can be efficiently and accurately identified from the financial network to form a target community. This process not only considers the direct connection between nodes, but more importantly, also considers the indirect and deep association, making the community division more reasonable and better meeting the specific needs of financial institutions in risk assessment, market analysis, customer service, etc., improving the accuracy and efficiency of business decision-making. At the same time, the setting of the preset overlap threshold provides flexibility, which can adjust the closeness of the community according to specific business scenarios and needs, enhancing the practicality and adaptability of the algorithm.

[0085] To improve the accuracy of determining the second-order neighbor coincidence degree between nodes, in the community determination method provided in Embodiment 1 of the present application, the second-order neighbor node set of the community node is determined, and the second-order neighbor node set of the neighbor node is determined; based on the second-order neighbor node set of the community node and the second-order neighbor node set of the neighbor node, the second-order neighbor coincidence degree between the community node and the neighbor node is determined.

[0086] In the financial network analysis in the embodiments of the present application, the community node refers to those nodes that have been determined to have business relevance, and these community nodes constitute the core of the initial community. The second-order neighbor node set of the community node can be determined first, that is, the nodes directly connected to the community node (first-order neighbors) and the nodes directly connected to these first-order neighbors (second-order neighbors) need to be found to form a set. Similarly, for the neighbor node (i.e., the node in the set of shell nodes connected to the initial community node), the second-order neighbor node set of the neighbor node also needs to be determined. In this way, the indirect connection between the community node and the neighbor node can be understood more deeply, because only considering the direct neighbors may not fully reflect the strength of the connection between the nodes and the complexity of the community structure.

[0087] In the embodiments of the present application, the second-order neighbor coincidence degree between the community node and the neighbor node can be calculated according to the second-order neighbor node set of the community node and the second-order neighbor node set of the neighbor node. The second-order neighbor coincidence degree is an index that measures the closeness of the indirect connection between nodes. By calculating the second-order neighbor coincidence degree, the connection between nodes indirectly established through first-order neighbors can be more accurately evaluated. This is crucial for determining which neighbor nodes should be absorbed into the community and how the community boundary should be dynamically adjusted.

[0088] In the embodiments, by introducing the second-order neighbor node set and the second-order neighbor coincidence degree, the application effect of the local community discovery algorithm in the financial network is improved. Not only can the shell nodes directly connected to the known important nodes (community nodes) be accurately identified, but also the indirect connection established through the first-order neighbors can be more comprehensively considered, so that the real correlation degree between nodes can be more accurately evaluated. It is especially suitable for the financial business scenario, because in the financial network, the connection between nodes is often not limited to direct transactions or relationships, but also may form more complex connections through indirect factors such as common transaction counterparties and directions of fund flow. By calculating and comparing the second-order neighbor coincidence degree, nodes that have deep connections with the community nodes can be more accurately identified, helping financial institutions more effectively perform business such as risk assessment, market segmentation, and customer relationship management, improving the accuracy and efficiency of decision-making.

[0089] In order to improve the accuracy of determining the second-order neighbor node set of the node, in the community determining method provided in Embodiment 1 of the present application, a first node set composed of all adjacent nodes of the node is determined, wherein the node is a community node or a neighbor node; a first-order neighbor node set is determined based on the node and the first node set; a second node set composed of adjacent nodes of each first node in the first node set is determined; and a second-order neighbor node set is determined based on the first-order neighbor node set and all second node sets.

[0090] In the embodiment of the present application, the first-order neighbor node set is defined as a node set composed of the node itself and all neighbors of the node, and in particular, the node itself is included. The definition is as follows using formula (2):

[0091] Γ 1 (x) = x U Γ(x) (2);

[0092] wherein Γ 1 (x) represents the first-order neighbor node set of node x, and Γ(x) represents a node set composed of all neighbors of node x.

[0093] The second-order neighbor node set is defined as a node set composed of the node itself and the first-order neighbor node set and the second-order neighbor of the node. The definition is as follows using formula (3):

[0094] Γ 2 (x) = x U Γ(x) U Γ(Γ(x)) (3);

[0095] wherein Γ 2 (x) represents the second-order neighbor node set of node x, and Γ(Γ(x)) represents a node set composed of the second-order neighbor of node x.

[0096] In the embodiment of the present application, the "first node set" refers to a set composed of all directly connected nodes (i.e., adjacent nodes) of a selected node (which can be a community node or a neighbor node). In this way, the direct social network environment of the node can be established, that is, at the first-order relationship level, which nodes are closely connected to the selected node. In the financial field, the selected node can be a known important user account, and the first node set contains all other accounts that have had direct transactions or contacts with this account.

[0097] Then, a first-order neighbor node set is determined based on the node and the first node set. Here, the first-order neighbor node set refers to the selected node itself plus all nodes in the first node set of the selected node, that is, the first-order neighbor node set is a set containing the selected node and its directly connected nodes. In this way, it is ensured that the selected node itself is also taken into account, because the attributes and behaviors of the node itself are also important for the influence on the community structure.

[0098] In the embodiments of the present application, the "second node set" refers to a set composed of the direct neighbors of each node in the first node set. The role of the second node set is to explore the second layer depth of the network structure, that is, to further expand the field of view through the first-order neighbors of the selected node and understand the wider indirect contact network. In the financial network, this can help financial institutions identify those accounts that have no direct transaction records with the selected node, but form indirect connections through common trading partners.

[0099] Then, based on the first-order neighbor node set and all second node sets, the second-order neighbor node set is determined. Here, the "second-order neighbor node set" refers to the set formed by merging the first-order neighbor node set and all second node sets and removing duplicate nodes, which can comprehensively present the network range touched by the selected node through the first-order neighbors and the second-order neighbors. In the financial business scenario, through the second-order neighbor node set, those accounts that have indirect connections with the selected node or its first-order neighbors can be identified. These accounts may be due to common investment preferences, similar trading patterns or the establishment of some association through a third party, and are crucial for building a more complete and detailed community structure diagram.

[0100] In the embodiments, by gradually constructing the first node set, the first-order neighbor node set and the second node set, the second-order neighbor node set is finally defined, and the deep-level connections between nodes in the financial network are comprehensively captured. This method can more accurately identify indirect connections between financial users, helping financial institutions make more accurate decisions in risk assessment, customer classification, market analysis and other business fields. Compared with the traditional method of analyzing only the first-order neighbors, the present embodiments can more completely depict the role and influence of nodes in the network, especially for those nodes that have an impact on core nodes through indirect means, making the application of community discovery algorithms in financial network analysis more effective and comprehensive. In addition, through the construction of the second-order neighbor node set, the algorithm can also detect potential community boundaries in the network, further optimizing the accuracy and efficiency of community division.

[0101] In order to further improve the accuracy of determining the second-order neighbor coincidence degree, in the community determination method provided in Embodiment 1 of the present application, the intersection set between the second-order neighbor node set of the community node and the second-order neighbor node set of the neighbor node is determined; the first node number of the second-order neighbor node set of the community node is determined, and the second node number of the second-order neighbor node set of the neighbor node is determined; based on the intersection node number of the intersection set, the first node number and the second node number, the second-order neighbor coincidence degree is determined.

[0102] In the embodiments of the present application, the second-order neighbor node coincidence degree is defined as the degree of coincidence of the second-order neighbor node sets of nodes x and y. It is described using formula (4) as follows:

[0103]

[0104] where Γ 2 (x) denotes the set of second-order neighbors of node x, Γ 2 (y) denotes the set of second-order neighbors of node y.

[0105] Specifically, the intersection set between the set of second-order neighbors of the community node and the set of second-order neighbors of the neighbor node can be determined first. Then, the total number of nodes in the set of second-order neighbors of the community node (i.e., the first number of nodes) and the total number of nodes in the set of second-order neighbors of the neighbor node (i.e., the second number of nodes) can be calculated. Subsequently, based on the number of nodes in the intersection set (i.e., the number of intersection nodes), the first number of nodes, and the second number of nodes, the "second-order neighbor overlap degree" can be calculated using formula (4). The second-order neighbor overlap degree is an index that measures how closely connected two nodes (the community node and the neighbor node) are in the second-order network. Formula (4) shows that the second-order neighbor overlap degree is the ratio of the number of nodes in the intersection set to the minimum number of the set of second-order neighbors of the community node and the set of second-order neighbors of the neighbor node. In this way, the strength of the indirect connection between the community node and the neighbor node can be quantified, which helps to identify which neighbor nodes should be absorbed into the community, thereby defining the boundary of the community. A higher second-order neighbor overlap degree means that the set of second-order neighbors of the two nodes has a larger overlap, indicating that the indirect connection between the two nodes in the network is close, and the likelihood of belonging to the same community is greater.

[0106] In this embodiment, by introducing the set of second-order neighbors and calculating the second-order neighbor overlap degree, the application accuracy and efficiency of the local community discovery algorithm in complex networks are improved. In financial network analysis, this method can help financial institutions more accurately identify nodes that have deep indirect connections with target community nodes, and then build a more rigorous and scientific community model. Compared with traditional community discovery algorithms based only on first-order neighbors, this embodiment can capture more complex and hidden structures in the network, improving the accuracy of community division and the effect of market segmentation. In addition, the calculation of the second-order neighbor overlap degree provides a quantitative standard for dynamically adjusting the community boundary, so that the algorithm can flexibly identify and expand the community according to actual business needs and network changes, ultimately achieving the purpose of optimizing business processes and enhancing decision support.

[0107] In the embodiment of the present application, the SNC algorithm is also applicable to relatively sparse networks. Figure 4 is a schematic diagram of a relatively sparse example network according to Embodiment 1 of the present application, as Figure 4 shown, there are a total of 11 nodes (0 to 10) and 13 edges.

[0108] with Figure 4The network shown is an example, and the SNC algorithm is used for community division. The first-order neighbor node set of node 0 is: Γ 1 (0) = {0, 1}, the second-order neighbor node set of node 0 is: Γ 2 (0) = {0, 1, 2, 5}. The first-order neighbor node set of node 1 is Γ 1 (1) = {0, 1, 2, 5}, the second-order neighbor node set of node 1 is Γ 2 (1) = {0, 1, 2, 3, 4, 5}. The second-order neighbor overlap degree C(0, 1) of node 0 and node 1 is calculated according to formula (5):

[0109]

[0110] Figure 5 is a schematic diagram of an example network with a second-order neighbor overlap degree according to the SNC algorithm of embodiment 1 of the present application, as Figure 5 shown, the second-order neighbor overlap degrees of nodes at both ends of all edges are calculated according to the SNC algorithm and marked on the example network, that is, the second-order neighbor overlap degree between node 0 and node 1 is 1; the second-order neighbor overlap degree between node 1 and node 5 is 1; the second-order neighbor overlap degree between node 1 and node 2 is 1; the second-order neighbor overlap degree between node 2 and node 3 is 0.83; the second-order neighbor overlap degree between node 5 and node 4 is 0.85; the second-order neighbor overlap degree between node 3 and node 4 is 1; the second-order neighbor overlap degree between node 4 and node 6 is 0.75; the second-order neighbor overlap degree between node 6 and node 10 is 1; the second-order neighbor overlap degree between node 6 and node 7 is 1; the second-order neighbor overlap degree between node 7 and node 10 is 1; the second-order neighbor overlap degree between node 10 and node 9 is 1; the second-order neighbor overlap degree between node 9 and node 8 is 1; the second-order neighbor overlap degree between node 7 and node 8 is 1.

[0111] Adjust the value of the parameter q in the SNC algorithm to the interval (0.75, 0.85], then Figure 5 the example network shown will be disconnected from nodes 4 and 6 into two small communities C1 and C2, C1 contains {0, 1, 2, 3, 4, 5}, and C2 contains {6, 7, 8, 9, 10}. In Figure 5 the example network, there is a node 0 with a degree (i.e., the number of connected edges) of 1, and a node 5 with a degree of 2 (node 5 does not exist in the graph structure), and these special nodes can still be correctly divided into communities under the SNC algorithm.

[0112] Figure 6 is a schematic diagram of the SNC algorithm process according to embodiment 1 of the present application, as Figure 6As shown, the known important node r can be input first, the second-order neighbor coincidence coefficient q can be calculated, then the neighbor coincidence C(x, y) of nodes at both ends of all edges can be calculated, whether the local community can be expanded according to the neighbor coincidence is judged, if the local community D can be expanded, the shell node set N is updated; if the local community cannot be expanded, the local community is returned.

[0113] In some optional embodiments, the effectiveness of the local community discovery algorithm can be analyzed by using three indexes of Precision, Recall and F-score (F value, i.e., harmonic mean of Precision and Recall). The real community to which the node i belongs is denoted as C i The node i is taken as a specific node, the local community obtained according to the local community discovery algorithm is denoted as D, and then the evaluation index is defined as follows:

[0114]

[0115] Wherein, |D∩C i | represents the number of nodes in the intersection between the local community D and the real community C i , |D| represents the number of nodes in the local community D, and |C i | represents the number of nodes in the real community C i .

[0116] In this embodiment, the SNC algorithm proposed in this embodiment can be compared with current community discovery algorithms (for example, Clauset algorithm (i.e., community discovery algorithm based on modularity maximization principle), LWP (Local Weakly Percolated) algorithm (i.e., local weakly percolated community algorithm), GMAC (Greedy Maximization of Compactness-Isolation) algorithm (i.e., greedy maximization of compactness-isolation algorithm), CNWNN (Common Neighbor Weighted Node Neighborhood) algorithm (i.e., common neighbor weighted node neighborhood algorithm) and FSR (Fuzzy Similarity Relation) algorithm (i.e., fuzzy similarity relation algorithm)), and it can be found that the SNC algorithm has better effect.

[0117] Figure 7 is a schematic diagram of experimental results of comparison of different algorithms according to Embodiment 1 of the present application on multiple real networks, as shown in Figure 7As shown in the figure, the SNC algorithm proposed in this embodiment is used together with the Clauset algorithm, LWP algorithm, GMAC algorithm, CNWNN algorithm, and FSR algorithm to perform community division on multiple real networks (such as the Karate network (a club network), the Dolphins network (a dolphin social network), the Polbooks network (a book network), and the Football network (a sports competitive relationship network)). Based on the division results of each algorithm, the Precision, Recall, and F-score indicators are calculated. Then, the experimental results of each real network are displayed in a bar graph. For example, Figure 7 (a) shows the experimental results of Clauset algorithm, LWP algorithm, GMAC algorithm, CNWNN algorithm, FSR algorithm, and SNC algorithm on Karate network, that is, the index values ​​of different algorithms in Precision, Recall and F-score (i.e. PRF.Value); Figure 7 (b) shows the experimental results of Clauset algorithm, LWP algorithm, GMAC algorithm, CNWNN algorithm, FSR algorithm, and SNC algorithm on Dolphins network, that is, the index values ​​of different algorithms on Precision, Recall and F-score (i.e. PRF.Value); Figure 7 (c) shows the experimental results of Clauset algorithm, LWP algorithm, GMAC algorithm, CNWNN algorithm, FSR algorithm, and SNC algorithm on Polbooks network, that is, the index values ​​of different algorithms on Precision, Recall and F-score (i.e. PRF.Value); Figure 7 (d) shows the experimental results of Clauset algorithm, LWP algorithm, GMAC algorithm, CNWNN algorithm, FSR algorithm, and SNC algorithm on the Football network, that is, the index values ​​of different algorithms in Precision, Recall and F-score (i.e. PRF.Value).

[0118] from Figure 7 It can be seen that the SNC algorithm of this embodiment has obvious advantages over other algorithms in the Karate, Dolphins, and Polbooks network data. The F-score in the Karate and Dolphins networks exceeds 0.9, and the advantage is particularly obvious on the Dolphins network. This is because there are many nodes with a degree of 1 in the Dolphins network, and the SNC algorithm can correctly classify nodes even if the node has a degree of 1.

[0119] Figure 8is a schematic diagram of experimental results of setting different q values on multiple real networks according to the SNC algorithm of embodiment 1 of the present application, as shown in Figure 8 , q values are set to 1, 0.95, 0.9, 0.85, 0.8, 0.75, 0.7 in karate, dolphins, polbooks and football respectively, and how the three evaluation indexes change under different q values is shown by a column chart. For example, Figure 8 (a) in shows the experimental results of different q values on the karate network, that is, the index values (i.e., P-R-F.Value) of Precision, Recall and F-score are calculated respectively after setting different q values; Figure 8 (b) in shows the experimental results of different q values on the dolphins network, that is, the index values (i.e., P-R-F.Value) of Precision, Recall and F-score are calculated respectively after setting different q values; Figure 8 (c) in shows the experimental results of different q values on the polbooks network, that is, the index values (i.e., P-R-F.Value) of Precision, Recall and F-score are calculated respectively after setting different q values; Figure 8 (d) in shows the experimental results of different q values on the football network, that is, the index values (i.e., P-R-F.Value) of Precision, Recall and F-score are calculated respectively after setting different q values.

[0120] As can be seen from Figure 8 , in the four networks, the Precision index is also decreasing as the q value decreases, indicating that only in the case of a large degree of overlap of second-order neighbor nodes, the local community mined is more accurate. The Recall index increases as the q value decreases, indicating that the smaller the limit of the degree of overlap of second-order neighbor nodes, the larger the local community mined, and the more accurately the local community where the node is covered. The F-score index basically has a trend of first increasing and then decreasing. Therefore, in actual application, the q value can be set according to this feature (i.e., the change of the F-score index value).

[0121] The community determination method provided by the embodiment of the present application can accurately divide the target community in the financial network by constructing the initial community and the shell node set thereof according to the determined initial node, using the preset second-order neighbor overlap algorithm, so that the user identifier related to the financial business can be effectively identified, the user indicated by the node of the target community is highly matched with the business demand, the interference of invalid users is avoided, the execution quality and safety of the financial business are improved, and the purpose of accurately positioning the target community in the financial network is achieved, thereby improving the technical effect of business efficiency.

[0122] It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described herein can be executed in an order different from that shown.

[0123] Embodiment 2

[0124] The embodiment of the present application also provides a community determining apparatus. It should be noted that the community determining apparatus of the embodiment of the present application can be used to execute the community determining method provided by the embodiment of the present application. The community determining apparatus provided by the embodiment of the present application is introduced as follows.

[0125] According to the embodiment of the present application, a device for implementing the above-mentioned community determining method is also provided. Figure 9 is a schematic diagram of an optional community determining apparatus according to the embodiment of the present application, as shown in the figure, the community determining apparatus can include a first determining unit 90, a constructing unit 91 and a second determining unit 92. Figure 9

[0126] The first determining unit 90 is configured to determine an initial node in a financial network based on a business purpose of a financial business, wherein the financial network is composed of a plurality of nodes, and each node indicates a user identifier;

[0127] The constructing unit 91 is configured to construct an initial community and a shell node set based on the initial node, wherein the initial community is a node set containing the initial node, and the shell node set includes nodes connected to the nodes in the initial community but not included in the initial community;

[0128] The second determining unit 92 is configured to determine a target community in the financial network based on the initial community and the shell node set by using a preset second-order neighbor coincidence degree algorithm, wherein each node in the target community indicates a user identifier required for processing the financial business.

[0129] The community determining apparatus provided by the embodiment of the present application can accurately divide the target community in the financial network by constructing the initial community and its shell node set according to the determined initial node and using the preset second-order neighbor coincidence degree algorithm, so that the user identifiers related to the financial business can be effectively identified, the nodes in the target community can be ensured to be highly matched with the business requirements, the interference of invalid users can be avoided, the execution quality and safety of the financial business can be improved, and the purpose of accurately positioning the target community in the financial network is achieved, thereby achieving the technical effect of improving the business efficiency.

[0130] ​Optionally, the first determining unit comprises: a first determining module, configured to determine at least one target user based on a service purpose of the financial service; and a second determining module, configured to determine a node corresponding to the target user in the financial network as the initial node.

[0131] Optionally, the constructing unit comprises: a first constructing module, configured to construct an initial community based on the initial node; a first querying module, configured to query neighbor nodes adjacent to the initial node in the financial network; and a second constructing module, configured to construct the set of shell nodes based on all the neighbor nodes.

[0132] Optionally, the second determining unit comprises: a first traversing module, configured to traverse the initial community and select a node in the initial community as a community node; a third determining module, configured to determine a second-order neighbor coincidence degree between the community node and any neighbor node of the community node; a first adding module, configured to add the neighbor node to the initial community and add all neighbor nodes of the neighbor node to the set of shell nodes, if the second-order neighbor coincidence degree is greater than or equal to a preset coincidence threshold and the neighbor node belongs to the set of shell nodes; and a fourth determining module, configured to continue to determine the second-order neighbor coincidence degree between the selected community node and any neighbor node of the community node until the set of shell nodes is empty or the second-order neighbor coincidence degree between any node in the set of shell nodes and the community node is less than the preset coincidence threshold, to obtain the target community.

[0133] Optionally, the third determining module comprises: a first determining submodule, configured to determine a second-order neighbor node set of the community node and a second-order neighbor node set of the neighbor node; and a second determining submodule, configured to determine the second-order neighbor coincidence degree between the community node and the neighbor node based on the second-order neighbor node set of the community node and the second-order neighbor node set of the neighbor node.

[0134] Optionally, the first determining submodule comprises: a third determining submodule, configured to determine a first node set composed of all neighbor nodes of a node, wherein the node is the community node or the neighbor node; a fourth determining submodule, configured to determine a first-order neighbor node set based on the node and the first node set; a fifth determining submodule, configured to determine a second node set composed of neighbor nodes of each first node in the first node set; and a sixth determining submodule, configured to determine the second-order neighbor node set based on the first-order neighbor node set and all the second node sets.

[0135] Optionally, the second determining sub-module comprises: a seventh determining sub-module, configured to determine a set of intersections between the second-order neighbor node set of the community node and the second-order neighbor node set of the neighbor node; an eighth determining sub-module, configured to determine a first number of nodes in the second-order neighbor node set of the community node and a second number of nodes in the second-order neighbor node set of the neighbor node; and a ninth determining sub-module, configured to determine the second-order neighbor overlap degree based on the number of intersection nodes in the set of intersections, the first number of nodes and the second number of nodes.

[0136] The community determining apparatus can further comprise a processor and a memory, and the first determining unit 90, the constructing unit 91, the second determining unit 92 and the like are stored in the memory as program units, and the corresponding functions are realized by the processor executing the program units stored in the memory.

[0137] The processor comprises a core, and the core retrieves the corresponding program units from the memory. The core can be one or more, and the core parameters are adjusted to determine the target community in the financial network based on the initial community and the shell node set by using the preset second-order neighbor overlap degree algorithm.

[0138] The memory can comprise a non-permanent memory in a computer readable medium, a random access memory (RAM) and / or a non-volatile memory such as a read-only memory (ROM) or a flash memory (flash RAM), and the memory comprises at least one memory chip.

[0139] It should be noted that the first determining unit 90, the constructing unit 91 and the second determining unit 92 correspond to steps S301 to S303 in Embodiment 1, and the units and the corresponding steps have the same instances and application scenarios, but are not limited to the content disclosed in Embodiment 1. It should be noted that the units can be hardware components or software components stored in the memory (for example, the memory 204) and processed by one or more processors (for example, the processors 202a, 202b, …, 202n), and the units can also be a part of the apparatus and can run in the computer terminal 20 provided in Embodiment 2.

[0140] Embodiment 3

[0141] Embodiments of the present application can provide an electronic device, Figure 10 is a structural block diagram of an electronic device according to an embodiment of the present application. As shown in the figure, the electronic device can comprise one or more (only one is shown in the figure) processors 1002, a memory 1004, a storage controller, and a peripheral interface, wherein the peripheral interface is connected with a radio frequency module, an audio module and a display. Figure 10 Figure 10

[0142] ​​The memory can be configured to store software programs and modules, such as program instructions / modules corresponding to the community determining method and apparatus in the embodiments of the present application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, that is, implements the community determining method described above. The memory can include a high-speed random access memory, and can further include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some examples, the memory can further include a memory remotely arranged with respect to the processor, and the remote memory can be connected to the terminal through a network. Examples of the network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0143] The processor can call information and application programs stored in the memory through the transmission device to perform the following steps: determining an initial node in a financial network based on a business purpose of a financial service, wherein the financial network is composed of a plurality of nodes, and each node indicates a user identifier; constructing an initial community and a shell node set based on the initial node, wherein the initial community is a node set containing the initial node, and the shell node set includes nodes connected to the nodes in the initial community but not included in the initial community; determining a target community in the financial network based on the initial community and the shell node set using a preset second-order neighbor coincidence degree algorithm, wherein each node in the target community indicates a user identifier required for processing the financial service.

[0144] The processor can call information and application programs stored in the memory through the transmission device to perform the following steps: determining at least one target user based on a business purpose of a financial service; and determining a node corresponding to a target user identifier of the target user in a financial network as an initial node.

[0145] The processor can call information and application programs stored in the memory through the transmission device to perform the following steps: constructing an initial community based on the initial node; querying neighbor nodes adjacent to the initial node in the financial network; and constructing a shell node set based on all neighbor nodes.

[0146] The processor can call information and application programs stored in the memory through the transmission device to perform the following steps: traversing the initial community and selecting a node in the initial community as a community node; determining a second neighbor coincidence degree between the community node and any neighbor node of the community node; in a case where the second neighbor coincidence degree is greater than or equal to a preset coincidence threshold and the neighbor node belongs to a shell node set, adding the neighbor node to the initial community and adding all adjacent nodes of the neighbor node to the shell node set; continuing to determine the second neighbor coincidence degree between the selected community node and any neighbor node of the community node until the shell node set is empty or the second neighbor coincidence degree between any node in the shell node set and the community node is all less than the preset coincidence threshold, to obtain a target community.

[0147] The processor can call information and application programs stored in the memory through the transmission device to perform the following steps: determining a second neighbor node set of the community node and determining a second neighbor node set of the neighbor node; determining a second neighbor coincidence degree between the community node and the neighbor node based on the second neighbor node set of the community node and the second neighbor node set of the neighbor node.

[0148] The processor can call information and application programs stored in the memory through the transmission device to perform the following steps: determining a first node set composed of all adjacent nodes of a node, wherein the node is a community node or a neighbor node; determining a first neighbor node set based on the node and the first node set; determining a second node set composed of adjacent nodes of each first node in the first node set; determining a second neighbor node set based on the first neighbor node set and all second node sets.

[0149] The processor can call information and application programs stored in the memory through the transmission device to perform the following steps: determining an intersection set between the second neighbor node set of the community node and the second neighbor node set of the neighbor node; determining a first node number of the second neighbor node set of the community node and determining a second node number of the second neighbor node set of the neighbor node; determining a second neighbor coincidence degree based on an intersection node number of the intersection set, the first node number and the second node number.

[0150] By constructing an initial community and a shell node set thereof according to the determined initial node, the preset second neighbor coincidence degree algorithm can be used to accurately divide a target community in a financial network, so that a user identifier related to a financial service can be effectively identified, the nodes of the target community are highly matched with service requirements, invalid users are avoided, the execution quality and safety of the financial service are improved, and the purpose of accurately positioning the target community in the financial network is achieved, thereby improving the technical effect of service efficiency.

[0151] It can be understood by those skilled in the art that Figure 10 The structure shown is for illustration only, and the electronic device may also be a terminal device such as a smart phone, a tablet computer, a PDA, or a mobile Internet device (MID). Figure 10 It does not limit the structure of the above electronic device. For example, the electronic device may also include Figure 10 More or fewer components (such as network interfaces, display devices, etc.) shown in, or with Figure 10 Different configurations shown.

[0152] A person skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing the hardware related to the terminal device through a program, and the program can be stored in a computer-readable storage medium, which may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0153] Example 4

[0154] The embodiment of the present application further provides a storage medium. Optionally, in this embodiment, the storage medium can be used to store the program code executed by the community determination method provided in the first embodiment.

[0155] Optionally, in this embodiment, the above-mentioned storage medium may be located in any computer terminal in a computer terminal group in a computer network, or in any mobile terminal in a mobile terminal group.

[0156] The present application also provides a computer program product, which, when executed on a data processing device, is adapted to perform the steps of the community determination method.

[0157] The serial numbers of the above embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0158] In the above embodiments of the present application, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0159] In several embodiments provided in the present application, it should be understood that the disclosed technology can be implemented by other ways. Among them, the above-described device embodiments are only schematic, for example, the division of the units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed units can be indirect coupling or communication connection through some interfaces, units or modules, and can be electrical or other forms.

[0160] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, i.e., they can be located in one place or distributed on a plurality of network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0161] In addition, the functional units in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0162] The integrated unit, if realized in the form of a software functional unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on this understanding, the technical solutions of the present application essentially or the part of the prior art that contributes to the technical solutions or the whole or part of the technical solutions can be embodied in the form of a software product, which is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The foregoing storage medium includes: a U disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a mobile hard disk, a magnetic disk or an optical disk, and various media that can store program codes.

[0163] The above is only the preferred embodiment of the present application, and it should be pointed out that for ordinary skilled persons in the art, without departing from the principles of the present application, a number of improvements and refinements can be made, and these improvements and refinements should be considered as the protection scope of the present application.

Claims

1. A community determination method, characterized in that: include: Determining an initial node in a financial network based on a business purpose of the financial service, wherein the financial network is composed of a plurality of nodes, each node indicating a user identifier; Based on the initial node, construct an initial community and a shell node set, wherein the initial community is a node set including the initial node, and the shell node set includes nodes connected to nodes in the initial community but not included in the initial community; Based on the initial community and the shell node set, a preset second-order neighbor coincidence algorithm is used to determine a target community in the financial network, wherein the user identifier indicated by each node in the target community is the user identifier required for processing the financial business.

2. The community determination method according to claim 1, characterized in that: Based on the business objectives of the financial business, the steps for determining the initial node in the financial network include: Identify at least one target user based on the business purpose of the financial service; The node corresponding to the target user identification of the target user in the financial network is determined as the initial node.

3. The community determination method according to claim 1, characterized in that: The steps of constructing an initial community and a shell node set based on the initial node include: Building the initial community based on the initial nodes; Querying neighbor nodes adjacent to the initial node in the financial network; The shell node set is constructed based on all the neighbor nodes.

4. The community determination method according to claim 1, characterized in that: The step of determining a target community in the financial network based on the initial community and the shell node set using a preset second-order neighbor overlap algorithm includes: Traversing the initial community, and selecting nodes in the initial community as community nodes; Determining a second-order neighbor overlap between the community node and any neighbor node of the community node; If the second-order neighbor overlap is greater than or equal to a preset overlap threshold and the neighbor node belongs to the shell node set, add the neighbor node to the initial community and add all adjacent nodes of the neighbor node to the shell node set; Continue to determine the second-order neighbor overlap between the selected community node and any neighbor node of the community node until the shell node set is empty or the second-order neighbor overlap between any node in the shell node set and the community node is less than the preset overlap threshold, thereby obtaining the target community.

5. The community determination method according to claim 4, characterized in that: The step of determining the second-order neighbor overlap between the community node and any neighbor node of the community node includes: Determining a set of second-order neighbor nodes of the community node, and determining a set of second-order neighbor nodes of the neighbor node; The second-order neighbor coincidence degree between the community node and the neighbor node is determined based on the second-order neighbor node set of the community node and the second-order neighbor node set of the neighbor node.

6. The community determination method according to claim 5, characterized in that: The steps of determining the set of second-order neighbor nodes of the community node and determining the set of second-order neighbor nodes of the neighbor node include: Determine a first node set consisting of all adjacent nodes of a node, wherein the node is the community node or the neighbor node; Determining a first-order neighbor node set based on the node and the first node set; Determine a second node set consisting of adjacent nodes of each first node in the first node set; The second-order neighbor node set is determined based on the first-order neighbor node set and all the second node sets.

7. The community determination method according to claim 5, characterized in that: The step of determining the second-order neighbor overlap between the community node and the neighbor node based on the second-order neighbor node set of the community node and the second-order neighbor node set of the neighbor node includes: Determine an intersection set between the set of second-order neighbor nodes of the community node and the set of second-order neighbor nodes of the neighbor node; Determining a first number of nodes in the second-order neighbor node set of the community node, and determining a second number of nodes in the second-order neighbor node set of the neighbor node; The second-order neighbor coincidence degree is determined based on the number of intersection nodes of the intersection set, the number of the first nodes, and the number of the second nodes.

8. A community determination device, characterized in that: include: A first determining unit is configured to determine an initial node in a financial network based on a business purpose of the financial business, wherein the financial network is composed of a plurality of nodes, each node indicating a user identifier; a construction unit configured to construct an initial community and a shell node set based on the initial node, wherein the initial community is a node set including the initial node, and the shell node set includes nodes connected to nodes in the initial community but not included in the initial community; A second determining unit is configured to determine a target community in the financial network based on the initial community and the shell node set by using a preset second-order neighbor coincidence algorithm, wherein a user identifier indicated by each node in the target community is a user identifier required for processing the financial business.

9. A computer program product, characterized in that The method comprises a non-volatile computer-readable storage medium storing a computer program, wherein the computer program implements the community determination method according to any one of claims 1 to 7 when executed by a processor.

10. An electronic device, characterized in that: The method comprises one or more processors and a memory, wherein the memory is used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the community determination method according to any one of claims 1 to 7.

Images