Social network-based insurance fraud gang identification method, system and device, and medium

Abstract

The invention discloses an insurance fraud gang identification method, system and device based on a social network and a medium, and relates to the technical field of computers, and the method comprises the steps: building a social network relation graph through employing an insurance service relation based on role information and target information in a newly-added claim settlement event; according to the social network relation graph, obtaining a traversal starting point used for constructing a closed-loop relation graph; constructing a plurality of closed-loop relation diagrams for the plurality of traversal starting points by using a first adjacent point acquisition mode based on a claim event relation; and by using a second adjacent point acquisition mode based on the claim event relationship, merging the plurality of closed-loop relationship diagrams to obtain a target closed-loop relationship diagram, the node information in the target closed-loop relationship diagram being used for representing a target insurance fraud gang. The method is suitable for efficient troubleshooting of fraud case recognition, and the fraud case recognition accuracy and efficiency are improved.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a method, system, device and medium for identifying insurance fraud gangs based on social networks. Background Art

[0002] In recent years, the insurance industry has continuously played the role of an economic shock absorber and social stabilizer, serving and protecting people's lives. With the innovative development of insurance products and services, insurance fraud methods have also been upgraded. The current insurance fraud risks show a trend of concealment, organization, and gang formation. Effectively identifying insurance fraud gangs has become an important proposition in the current insurance anti-fraud.

[0003] At present, the use of data mining for fraud gang identification in China is mainly divided into two categories: one is the regular rule method, and the other is the machine learning method (generating a fraud identification model through training of fraud and non-fraud samples). Both methods have certain limitations. Among them, the regular rule method refers to analyzing existing fraud cases and manually extracting key features. If new cases trigger existing case features, they will be identified. The shortcomings of this method are that it is lagging and cannot identify new fraud cases. Second, criminals will also learn to avoid key features, making the effectiveness of the rules short-lived; the machine learning method refers to the use of algorithms of deep neural networks such as CNN and LSTM, and training models to generate fraud identification models by marking fraud and non-fraud cases. It is used for the detection and identification of new cases. The shortcomings of this method are that the sample size of fraud cases is small, resulting in an inappropriate ratio of fraud cases and non-fraud cases in the training samples, resulting in insufficient accuracy of the model in identifying fraud cases. Summary of the invention

[0004] In view of this, the present application provides a method, system, device and medium for identifying insurance fraud gangs based on social networks. The main purpose is to solve the technical problems that the existing regular rule-based method can only identify fraud cases with existing key features, the solution has low timeliness, and the machine learning method has high requirements for sample data, so the accuracy of fraud case identification is low.

[0005] According to one aspect of the present application, a method for identifying insurance fraud gangs based on social networks is provided, the method comprising:

[0006] Based on the role information and target information in the newly added claims events, the social network relationship diagram is constructed using the insurance business relationship;

[0007] According to the social network relationship graph, obtaining a traversal starting point for constructing a closed-loop relationship graph;

[0008] Using the first neighbor point acquisition method based on the claim event relationship, construct multiple closed-loop relationship graphs for multiple traversal starting points;

[0009] By utilizing the second neighbor point acquisition method based on the claim event relationship, the target closed-loop relationship graph is obtained by merging the multiple closed-loop relationship graphs, and the node information in the target closed-loop relationship graph is used to characterize the target insurance fraud gang.

[0010] According to another aspect of the present application, a social network-based insurance fraud gang identification system is provided, the system comprising:

[0011] The first construction module is used to construct a social network relationship diagram using insurance business relationships based on role information and target information in the newly added claim event;

[0012] A traversal module, used to obtain a traversal starting point for constructing a closed-loop relationship graph according to the social network relationship graph;

[0013] A second construction module is used to construct multiple closed-loop relationship graphs for multiple traversal starting points by using a first neighbor point acquisition method based on a claim event relationship;

[0014] A merging module is used to utilize a second neighbor point acquisition method based on the claim event relationship to merge the multiple closed-loop relationship graphs to obtain a target closed-loop relationship graph, wherein the node information in the target closed-loop relationship graph is used to characterize the target insurance fraud gang.

[0015] According to another aspect of the present application, a computer storage medium is provided, on which a computer program is stored, and when the program is executed by a processor, the above-mentioned insurance fraud gang identification method based on social network is implemented.

[0016] According to another aspect of the present application, a computer device is provided, including a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, wherein the processor implements the above-mentioned social network-based insurance fraud gang identification method when executing the program.

[0017] Through the above-mentioned technical scheme, the social network-based insurance fraud gang identification method, system, device and medium provided by the present application are compared with the existing regular rule-based method that can only identify fraud cases with existing key features and has low timeliness, and the technical scheme based on machine learning method has high requirements for sample data and therefore has low accuracy in identifying fraud cases. The present application is based on the role information and target information in the newly added claims event, and uses insurance business relationships to construct a social network relationship graph; based on the social network relationship graph, a traversal starting point for constructing a closed-loop relationship graph is obtained; a first neighbor point acquisition method based on the claims event relationship is used to construct multiple closed-loop relationship graphs for multiple traversal starting points; a second neighbor point acquisition method based on the claims event relationship is used to merge the multiple closed-loop relationship graphs to obtain a target closed-loop relationship graph, and the node information in the target closed-loop relationship graph is used to characterize the target insurance fraud gang. It can be seen that the social network relationship graph constructed based on the role information and target information in the newly added claims event, through traversal analysis of each node in the social network relationship graph, and obtaining the correlation between each node based on the claims event relationship, thereby obtaining the target closed-loop relationship graph, that is, finding the target insurance fraud gang information through the target closed-loop relationship graph, which can not only avoid the limitations of the historical key features in the regular rule method, but also ensure the accuracy of identifying fraud cases.

[0018] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0020] Figure 1 A schematic diagram of a process of identifying an insurance fraud gang based on a social network provided in an embodiment of the present application is shown;

[0021] Figure 2 A schematic diagram of a process of another insurance fraud gang identification method based on a social network provided in an embodiment of the present application is shown;

[0022] Figure 3 A schematic diagram showing a target closed-loop relationship diagram of another insurance fraud gang identification method based on a social network provided in an embodiment of the present application;

[0023] Figure 4A schematic diagram of the structure of a social network-based insurance fraud gang identification system provided in an embodiment of the present application is shown;

[0024] Figure 5 A schematic structural diagram of another social network-based insurance fraud gang identification system provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0025] The present application will be described in detail below with reference to the accompanying drawings and in combination with embodiments. It should be noted that the embodiments and features in the embodiments of the present application can be combined with each other without conflict.

[0026] The existing regular rule-based methods can only identify fraud cases with existing key features, and the timeliness of the solutions is low. In addition, the machine learning methods have high requirements for sample data, so there are technical problems such as low accuracy in identifying fraud cases. This embodiment provides a method for identifying insurance fraud gangs based on social networks, which can analyze the correlation between each claim event based on the newly added claim event, so as to obtain a target closed-loop relationship graph, that is, to find the target insurance fraud gang information through the target closed-loop relationship graph, which can not only avoid the limitations of historical key features in the regular rule method, but also ensure the accuracy of fraud case identification. Figure 1 As shown, the above method comprises the following steps:

[0027] Step 101: Based on the role information and target information in the newly added claim settlement event, a social network relationship diagram is constructed using insurance business relationships.

[0028] In this embodiment, the traditional social network is a collection of social actors and their relationships. According to the needs of the actual application scenario, the social network in this embodiment is a collection of social actors with insurance business relationships and their related insurance subject information. Specifically, nodes are created based on role information and vehicle information in subject information, and the role information in the newly added claim event and the vehicle information in the subject information are used as the traversal starting point for constructing the social network relationship graph. By associating other newly added claim events, as well as nodes in the newly added claim event and historical claim events, a social network relationship graph for the newly added claim event is obtained, which can not only achieve further judgment on historical suspected fraud cases, but also mine fraud cases in newly added claim events.

[0029] Among them, different role information in the newly added claim event is obtained. Role information may include driver, car owner, third party (policyholder, insured, reporter, beneficiary, payee, insured, reporter, surveyor, etc.), target information may include vehicle, claim amount of claim event, etc., and the insurance business relationship may include the relationship between policyholder and insured (policy relationship), stakeholder relationship in the claim event (insured and payee relationship, beneficiary and insured relationship), other related person relationships in the claim event (reporter, surveyor), and other relationships related to insurance business. According to the needs of the actual application scenario, role information may also include enterprise information associated with personnel, that is, the unified social credit code of the enterprise associated with personnel as a unique identifier.

[0030] Step 102: According to the social network relationship graph, a traversal starting point for constructing a closed-loop relationship graph is obtained.

[0031] In this embodiment, in order to ensure the completeness of the constructed closed-loop relationship graph, each node in the social network relationship graph is traversed, and the node corresponding to the maximum number of claim events is used as the traversal starting point for constructing the closed-loop relationship graph; or, the node corresponding to the maximum number of claim events and the maximum value of claim amount information is used as the traversal starting point for constructing the closed-loop relationship graph. The maximum value of claim amount information refers to the sum of the compensation amounts of each claim event in the node, and the sum of the compensation amounts is the maximum value in each node of the social network relationship graph.

[0032] Step 103: construct multiple closed-loop relationship graphs for multiple traversal starting points by using the first neighbor point acquisition method based on the claim event relationship.

[0033] In this embodiment, after the traversal starting point is determined, by traversing other nodes, a node that has the same insurance business relationship with the traversal starting point is obtained and used as the first neighboring point of the traversal starting point. After determining the number of claim events stored in the node and the total amount of compensation for each claim event, continue to traverse other nodes that have the same insurance business relationship with the traversal starting point. If the number of claim events stored in the node is greater than the first neighboring point, or the number of claim events is the same but the total amount of compensation for each claim event is greater than the first neighboring point, then the first neighboring point is updated until the traversal is completed, the target first neighboring point is determined, and so on, a relationship chain is obtained, and then multiple closed-loop relationship graphs of the current traversal starting point are constructed based on the nodes in the relationship chain.

[0034] Step 104: Utilize the second neighbor point acquisition method based on the claim event relationship to merge the multiple closed-loop relationship graphs to obtain a target closed-loop relationship graph.

[0035] In this embodiment, for multiple closed-loop relationship graphs constructed from different traversal starting points, each closed-loop relationship graph is analyzed to see whether there is an associated node with the same claim event relationship, and based on the associated node in at least two closed-loop relationship graphs, at least two closed-loop relationship graphs are merged to obtain a target closed-loop relationship graph. The role information stored in each node in the target closed-loop relationship graph is used to characterize the target insurance fraud gang.

[0036] For this embodiment, according to the above scheme, based on the role information and target information in the newly added claim event, a social network relationship graph can be constructed using the insurance business relationship; according to the social network relationship graph, a traversal starting point for constructing a closed-loop relationship graph is obtained; a first neighbor point acquisition method based on the claim event relationship is used to construct multiple closed-loop relationship graphs for multiple traversal starting points; a second neighbor point acquisition method based on the claim event relationship is used to merge the multiple closed-loop relationship graphs to obtain a target closed-loop relationship graph, and the node information in the target closed-loop relationship graph is used to characterize the target insurance fraud gang, which is different from the existing regular rule-based method that can only Compared with the technical solutions based on machine learning methods that have high requirements for sample data and therefore have low accuracy in identifying fraud cases with existing key features, this embodiment constructs a social network relationship graph based on the role information and target information in the newly added claims event, traverses and analyzes each node in the social network relationship graph, and finds the relationship between different insurance roles and insurance targets in the insurance underwriting claims event based on the claims event relationship, thereby obtaining a target closed-loop relationship graph, that is, finding the target insurance fraud gang information through the target closed-loop relationship graph, effectively improving the accuracy and timeliness of identifying insurance fraud gangs.

[0037] Furthermore, as a refinement and expansion of the specific implementation of the above embodiment, in order to fully illustrate the specific implementation process of this embodiment, another insurance fraud gang identification method based on social networks is provided. This embodiment can be applied to the auto insurance system. This embodiment can find the association between different insurance roles and insurance objects in insurance underwriting claims events based on the relationship between claims events, find the target insurance fraud gang information, and effectively improve the accuracy and timeliness of identifying insurance fraud gangs. Figure 2 As shown, the method includes:

[0038] Step 201: Based on the role information and target information in the newly added claim settlement event, a social network relationship diagram is constructed using insurance business relationships.

[0039] During implementation, in order to illustrate the specific implementation method of step 201, as a preferred embodiment, step 201 includes: using the role information in the newly added claim event and the vehicle information in the target information as nodes in the social network relationship graph; for each node, obtaining the number of claim events associated with the node, and storing the number of claim events on the node; connecting the nodes using insurance business relationships to construct a social network relationship graph.

[0040] During implementation, the traversal can uniquely identify all roles with insurance business relationships (which may also include role-associated companies) and information on the subject matter, and clean up invalid information. Specifically, the identity card numbers of all insurance roles in the insurance business relationships (which may also include the unified social credit codes of the companies associated with the personnel) and the vehicle frame numbers are obtained as unique identifiers, and saved in a collection as nodes in the social network relationship graph. Each element identifier in the collection (corresponding to a node in the social network relationship graph) is unique. Each traversal extracts a new element identifier and checks whether the element identifier exists in the collection. If not, the new element identifier is added to the collection. If so, the element identifier is discarded to ensure the uniqueness of the element identifiers in the collection. Assuming that n element identifiers are extracted, the final collection is P = {p 1 ,p 2 ,…,p n}.

[0041] During implementation, the social network relationship graph includes: a first social network relationship graph constructed by traversing the newly added claim events, the first social network relationship graph includes the newly added claim events; or / and, a second social network relationship graph constructed by traversing historical claim events and the newly added claim events, the second social network relationship graph includes the newly added claim events and historical claim events.

[0042] This embodiment is different from the prior art in that it does not rely on specific restrictions and does not require searching for samples in advance for training. It is based on the characteristics of insurance fraud cases involving gangs, namely, the number of times that role-vehicle associations appear is significantly higher than normal, and the probability of occurrence significantly deviates from the laws of nature. Regular traversal is performed to find the associations between all insurance roles and objects. The traversal method includes traversing the role information and object information in newly added claims events to identify new types of fraud cases, and traversing newly added claims events and historical claims events based on the role information and object information in the newly added claims events to further determine historical suspected fraud cases and achieve the comprehensiveness of fraud case identification.

[0043] Step 202: According to the social network relationship graph, a traversal starting point for constructing a closed-loop relationship graph is obtained.

[0044] During implementation, in order to illustrate the specific implementation method of step 202, as a preferred embodiment, step 202 includes: obtaining the node corresponding to the maximum number of claim events in the social network relationship graph, and using it as the traversal starting point for constructing a closed-loop relationship graph; or, the target information includes the claim amount corresponding to each claim event, obtaining multiple nodes corresponding to the maximum number of claim events in the social network relationship graph, and determining the traversal starting point for constructing a closed-loop relationship graph based on the claim amount information in the multiple nodes.

[0045] In practice, the graphical representation of social networks is called social network relationship graph and social network relationship matrix. The social network relationship graph is a set of multiple nodes (social actors with insurance relationships) and links between nodes (social actors with insurance interest relationships). The nodes in the social network relationship graph are used to represent the roles (social actors) and targets with insurance business relationships, and the edges are used to represent various insurance business relationships between roles and between roles and targets, such as the relationship between the insured and the insured, the relationship between the main vehicle driver and the third party in the accident, etc. The relationship referred to is undirected. Among them, the social network relationship graph consists of a set of nodes N = {n 1 ,n 2 ,…,n k} and the connection line between nodes L={l 1 ,l 2 ,…,l m} is composed of.

[0046] According to the needs of the actual application scenario, starting from a certain element identifier, if the element identifier p n If a claim event is involved, the number is recorded as 1, and the number of claim events involved is n. After traversing each element identifier, a claim event number set is formed. in, Refers to the element identifier p n The number of claims involved, denoted by p 1 = max(n), that is, the element identifier involving the largest number of claims events is used as the traversal starting point P 1 Furthermore, if there are element identifiers with the same number of claim events, the element identifier with the largest claim amount is taken as P. 1 , if the two judgment conditions are the same, then randomly select an element identifier as the starting point P 1 .

[0047] Step 203: Utilize the first neighbor point acquisition method based on the claim event relationship to construct a relationship chain based on the current traversal starting point.

[0048] During implementation, the first neighboring point acquisition method is to acquire multiple nodes that have the same insurance business relationship with the traversal starting point, and determine the direct neighboring points of the traversal starting point according to the number of claims events corresponding to the same insurance business relationship; the direct neighboring point is that there are no other nodes between the traversal starting point and the neighboring point, and the indirect neighboring point is that there are other nodes between the traversal starting point and the neighboring point.

[0049] In the implementation, starting from the traversal starting point, if it is p 1 , traverse the element identifier p that has a claim event relationship with it i , the claim event relationship between the two is recorded as l 1,i , continue with p i Start by traversing the element identifier p that has a claim event relationship with it i+1 , at this time p 1 With p i+1 The relationship is l 1,i,i+1 , continue the above search process until the relationship no longer changes, thus obtaining a relationship chain L with a unique traversal starting point and different relationship lengths. The complete relationship chain is recorded as l 1,2,…,n It should be noted that if p 1 and p 2 There is a claim event relationship, denoted as l 12 , the relationship between claims events does not have directionality, that is, l 12 = l 21 For the convenience of description, the naming format of small number first and large number last is adopted.

[0050] Specifically, with the element identifier p i For example, select the element identifier with the largest number of claim events, or select the element identifier with the largest amount of compensation from the element identifiers with the same number of claim events. If the above two judgment conditions are the same, randomly select one from the element identifiers that meet the judgment conditions as p i , the claim event relationship between the two is recorded as l 1,i Similarly, continue with the element identifier p i As the starting point, the above process is carried out, and the associated element identifiers are no longer involved in the traversal process until no new relationship chains appear. Further, according to the above traversal order, the relationship chains are sorted, and the complete relationship chain set is recorded as l 1,2,…,n , including the first traversal of the chain denoted by The nth traversal of the chain is recorded as l (1,i,i+1)nAll nodes in the relationship chain set correspond to the nodes in the traversal chain. According to the needs of the actual application scenario, it can be determined that the fraud risk of the element identifier on the first traversal chain is the greatest, and the first traversal chain is used as the target relationship chain, or the traversal chain with the most element identifiers is used as the target relationship chain. The target relationship chain is not specifically limited here.

[0051] Step 204: Obtain indirect neighboring points in the relationship chain that have the same insurance business relationship with the traversal starting point, and construct multiple closed-loop relationship graphs for the current traversal starting point based on the indirect neighboring points.

[0052] Step 205: Utilize the second neighbor point acquisition method based on the claim event relationship to merge the multiple closed-loop relationship graphs to obtain a target closed-loop relationship graph.

[0053] During implementation, the node information in the target closed-loop relationship graph is used to characterize the target insurance fraud gang. If there are associated nodes with the same claim event relationship in at least two of the multiple closed-loop relationship graphs, the at least two closed-loop relationship graphs are merged based on the associated nodes to obtain the target closed-loop relationship graph. The associated nodes refer to nodes with the same claim event relationship that are not in the same closed-loop relationship graph.

[0054] In practice, in a social network relationship graph, if any node has a path connected to it, then the social network relationship graph is a closed-loop graph relationship, that is, by obtaining multiple nodes that have the same insurance business relationship with the traversal starting point, and taking the node corresponding to the maximum number of claims events as the first neighbor of the traversal starting point, and continuing to obtain the second neighbor of the first neighbor, until multiple closed-loop relationship graphs for the traversal starting point are obtained. Therefore, in the complete relationship chain set l 1,2,…,n In the example, if n and 1 have an insurance business relationship, it is considered that a closed loop can be formed. For example, if n=3, it is a 3-element ring, and if n=4, it is a 4-element ring. For example, based on the insured a, through the insurance business relationship (underwriting relationship, claim relationship, etc.), the shortest path is used to associate with other roles, thereby obtaining a closed-loop relationship graph. According to the needs of the actual application scenario, when a larger ring contains a smaller ring, the element in the small ring is considered to be one of the elements of the larger ring, and the element on the small ring is determined to be a high-risk fraud person / high-risk fraud enterprise / high-risk fraud vehicle / high-risk fraud insurance subject, thereby effectively reducing the time to obtain the closed-loop relationship graph.

[0055] Furthermore, for two or more different closed-loop relationship graphs, if at least one element between each two closed-loop relationship graphs is associated, the two or more closed-loop relationship graphs are merged to obtain a target closed-loop relationship graph for characterizing the target insurance fraud gang. Specifically, by calculating the degree of association between the elements on the ring and all the elements on the ring, multiple elements with the highest degree of association are determined to have the highest fraud risk, as shown in steps 206-208, and the determination method is not specifically limited here.

[0056] In a social network relationship diagram, when two nodes are connected by a straight line, the two nodes are called adjacent, and the node adjacent to a node is called the neighbor of the node. Therefore, the association degree can also be expressed as the n of a node. i The number of neighbors of i ). According to the needs of actual application scenarios, for multiple social network relationship graphs, the correlation degree can also be density, that is, the ratio of the number of actual association relationships in the social network relationship graph to the maximum number of possible relationships. If the density of a social network relationship graph is 1, it means that each node in the social network relationship graph is connected to other nodes. Conversely, if the density of the social network relationship graph is 0, it means that no nodes in the social network relationship graph are connected.

[0057] It should be noted that, based on the newly added elements and the elements not included in the closed loop, a new element set is formed. First, step 203 can be repeated to form a new relationship chain; second, step 204 can be repeated to find a new closed loop relationship diagram; third, step 205 can be repeated to find a new closed loop relationship diagram (for example Figure 3 The relationship chain 5-3-4, 0-1-2) in the target closed-loop relationship diagram is merged with the existing target closed-loop relationship diagram to update the target closed-loop relationship diagram; fourth, for the chain that is not included in the target closed-loop relationship diagram, by traversing the relationship between the elements in the above chain and the elements of the existing target closed-loop relationship diagram, if there is a situation where the associated element is greater than or equal to 2 (for example Figure 3 The relationship chain 5-3-4 in the target closed-loop relationship diagram is merged and processed to update the target closed-loop relationship diagram, and step 205 is repeated until the target closed-loop relationship diagram no longer changes; fifth, for elements that are not included in the target closed-loop relationship diagram, by traversing the relationship between the above elements and the existing target closed-loop relationship diagram elements, if there is a situation where the associated element is greater than or equal to 2 (for example Figure 3 The relationship chains 0-1-2 in the target closed-loop relationship graph are merged and processed to update the target closed-loop relationship graph, and step 205 is repeated until the target closed-loop relationship graph no longer changes.

[0058] Step 206: Determine the shortest distance between any two nodes by traversing any two nodes in the target closed-loop relationship graph, and construct an adjacency matrix based on the shortest distance.

[0059] Step 207: Calculate the sum of the shortest distances of each node in the target closed-loop relationship graph according to the adjacency matrix.

[0060] Step 208: The node with the shortest distance and satisfying the set conditions is regarded as a key member of the target insurance fraud gang.

[0061] In practice, the shortest distance is the shortest path distance, the shortest path between two nodes is the shortest path between the two nodes to achieve association, and the shortest path distance between two nodes is the length of the shortest path between the two nodes. Figure 3 As shown in the target closed loop relationship diagram h 1 For example, it contains a loop relationship point (0-5-4-6), and nodes (elements) 1, 2, and 3 are connected to the nodes on the loop relationship respectively. Starting from any vertex, by calculating the shortest distance of all nodes (other nodes) in the target closed-loop relationship graph through the vertex, the shortest distance sum is obtained, and at least one node with the shortest distance sum is taken as the key personnel of the gang. The details are as follows:

[0062] Step 2081, assuming that the distance between two nodes is 1, then the distance between the two nodes is the shortest path distance. Using the breadth-first traversal method, it can be seen that there are three paths between node 0 and node 4: 0-6-4, with a distance of 2; 0-5-3-4, with a distance of 3; 0-5-4, with a distance of 2. Then the shortest path distance between node 0 and node 4 is 2.

[0063] Step 2082, taking different nodes in the target closed-loop relationship graph as vertices, calculate the shortest distances to all nodes (other nodes), as shown in Table 1, and obtain the adjacency matrix by traversing the shortest distances from any node to all nodes (other nodes), as shown in Table 2, and sum the adjacency matrix by column to obtain the sum of the shortest distances of each node in the target closed-loop relationship graph.

[0064] Table 1 Traversal calculation of the shortest distance and

[0065] Point 0 Point 1 Point 2 Point 3 Point 4 Point 5 Point 6 Point 0 0 1 1 2 2 1 1 Point 1 1 0 1 3 3 2 2 Point 2 1 1 0 3 3 2 2 Point 3 2 3 3 0 1 1 2 Point 4 3 4 3 1 0 1 1 Point 5 1 2 2 1 1 0 2 Point 6 1 2 2 2 1 2 0

[0066] Table 2 The shortest distances and

[0067] Point 0 Point 1 Point 2 Point 3 Point 4 Point 5 Point 6 The shortest distance and 9 13 12 12 11 9 10

[0068] It can be seen that based on the shortest distance and sorting, the criticality of each node is ranked as node 0 (node ​​5) > node 6 > node 4 > node 2 (node ​​3) > node 1. Based on the social scenario, people on the nodes with the shortest distance and the smallest size can contact other people with less contact cost. The people on the nodes with the shortest distance and the smallest size have more important positions in the social network. Therefore, nodes 0 and 5 are considered to be key members of the target insurance fraud gang.

[0069] According to the needs of actual application scenarios, the breadth-first traversal method can use an auxiliary queue, enqueue any node as a vertex, mark it as visited, and then loop to check whether the queue is empty. If the queue is not empty, select the first element in the queue, enqueue all unvisited nodes associated with the element, and mark these nodes as visited; if the queue is empty, it means that all nodes have been traversed in a breadth-first manner.

[0070] By applying the technical solution of this embodiment, based on the role information and target information in the newly added claim event, a social network relationship graph is constructed using the insurance business relationship; according to the social network relationship graph, a traversal starting point for constructing a closed-loop relationship graph is obtained; using the first neighbor point acquisition method based on the claim event relationship, multiple closed-loop relationship graphs for multiple traversal starting points are constructed; using the second neighbor point acquisition method based on the claim event relationship, the multiple closed-loop relationship graphs are merged to obtain the target closed-loop relationship graph, so as to obtain a technical solution for finding a fraud gang, and further find key members by calculating the shortest path. Compared with the existing regular rule-based method that can only identify fraud cases with existing key features, the solution has low timeliness, and the technical solution based on the machine learning method has a low accuracy in identifying fraud cases due to the high requirements for sample data, it can be seen that by regularly traversing and finding the closed-loop relationship of the association relationship between different insurance roles and insurance targets in insurance underwriting claim events under the shortest path, the key members of the target insurance fraud gang can be effectively found.

[0071] Further, as Figure 1 The specific implementation of the method, the embodiment of the present application provides an insurance fraud gang identification system based on social network, such as Figure 4 As shown, the system includes: a first construction module 41, a traversal module 42, a second construction module 43, and a merging module 44.

[0072] The first construction module 41 is used to construct a social network relationship diagram using insurance business relationships based on role information and target information in the newly added claim settlement event.

[0073] The traversal module 42 is used to obtain a traversal starting point for constructing a closed-loop relationship graph according to the social network relationship graph.

[0074] The second construction module 43 is used to construct multiple closed-loop relationship graphs for multiple traversal starting points by using the first neighbor point acquisition method based on the claim event relationship.

[0075] The merging module 44 is used to utilize the second neighbor point acquisition method based on the claim event relationship to merge the multiple closed-loop relationship graphs to obtain a target closed-loop relationship graph, and the node information in the target closed-loop relationship graph is used to characterize the target insurance fraud gang.

[0076] In specific application scenarios, such as Figure 5 As shown, the first building module 41 includes a setting unit 411 , a storage unit 412 , and a connecting unit 413 .

[0077] The setting unit 411 is used to use the role information in the newly added claim settlement event and the vehicle information in the target information as nodes in the social network relationship graph.

[0078] The storage unit 412 is used to obtain, for each node, the number of claim settlement events associated with the node, and store the number of claim settlement events on the node.

[0079] The connection unit 413 is used to connect the nodes using insurance business relationships to construct a social network relationship graph.

[0080] In a specific application scenario, the social network relationship graph includes: a first social network relationship graph constructed by traversing the newly added claim events, the first social network relationship graph including the newly added claim events; or / and, a second social network relationship graph constructed by traversing historical claim events and the newly added claim events, the second social network relationship graph including the newly added claim events and historical claim events.

[0081] In a specific application scenario, the target information includes the claim amount corresponding to each claim event, and the traversal module 42 is specifically used to obtain the node corresponding to the maximum number of claim events in the social network relationship graph, and use it as the traversal starting point for constructing a closed-loop relationship graph; or, obtain multiple nodes corresponding to the maximum number of claim events in the social network relationship graph, and determine the traversal starting point for constructing a closed-loop relationship graph based on the claim amount information in the multiple nodes.

[0082] In a specific application scenario, the second building module 43 includes: a first acquiring unit 431 and a second acquiring unit 432 .

[0083] The first acquisition unit 431 is used to construct a relationship chain based on the current traversal starting point by using the first neighbor point acquisition method based on the claim event relationship; wherein the first neighbor point acquisition method is to acquire multiple nodes that have the same insurance business relationship with the traversal starting point, and determine the direct neighbor points of the traversal starting point according to the number of claim events corresponding to the same insurance business relationship; the direct neighbor points are nodes that do not exist between the traversal starting point and the neighbor points, and the indirect neighbor points are nodes that exist between the traversal starting point and the neighbor points.

[0084] The second acquisition unit 432 is used to acquire indirect neighboring points in the relationship chain that have the same insurance business relationship with the traversal starting point, and construct multiple closed-loop relationship graphs for the current traversal starting point based on the indirect neighboring points.

[0085] In a specific application scenario, the merging module 44 is specifically used to merge the at least two closed-loop relationship graphs based on the associated nodes to obtain a target closed-loop relationship graph if there are associated nodes with the same claim event relationship in at least two closed-loop relationship graphs among the multiple closed-loop relationship graphs.

[0086] In a specific application scenario, the system further includes: a determination module 45 .

[0087] The determination module 45 is used to determine the shortest distance between any two nodes in the target closed-loop relationship graph by traversing the any two nodes, and to construct an adjacency matrix based on the shortest distance; and, based on the adjacency matrix, to calculate the sum of the shortest distances of each node in the target closed-loop relationship graph; and, to use the nodes with the shortest distance and meeting the set conditions as the key members of the target insurance fraud gang.

[0088] It should be noted that for other corresponding descriptions of the functional units involved in the social network-based insurance fraud gang identification system provided in the embodiment of the present application, reference can be made to Figure 1 and Figure 2 The corresponding description in will not be repeated here.

[0089] Based on the above Figure 1 and Figure 2 The method shown in the embodiment of the present application accordingly provides a computer storage medium on which a computer program is stored, and when the program is executed by a processor, the above-mentioned Figure 1 and Figure 2 The social network-based insurance fraud gang identification method shown.

[0090] Based on this understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each implementation scenario of the present application.

[0091] Based on the above Figure 1 , Figure 2 The method shown, and Figure 4 , Figure 5 In order to achieve the above-mentioned purpose, the embodiment of the present application further provides a computer device, which can be a personal computer, a server, a network device, etc. The physical device includes a storage medium and a processor; the storage medium is used to store a computer program; the processor is used to execute the computer program to achieve the above-mentioned Figure 1 and Figure 2 The social network-based insurance fraud gang identification method shown.

[0092] Optionally, the computer device may further include a user interface, a network interface, a camera, a radio frequency (RF) circuit, a sensor, an audio circuit, a WI-FI module, etc. The user interface may include a display, an input unit such as a keyboard, etc., and the optional user interface may also include a USB interface, a card reader interface, etc. The network interface may optionally include a standard wired interface, a wireless interface (such as a Bluetooth interface, a WI-FI interface), etc.

[0093] Those skilled in the art will appreciate that the computer device structure provided in this embodiment does not constitute a limitation on the physical device, and may include more or fewer components, or a combination of certain components, or different component arrangements.

[0094] The storage medium may also include an operating system and a network communication module. The operating system is a program that manages the hardware and software resources of a computer device and supports the operation of information processing programs and other software and / or programs. The network communication module is used to achieve communication between components within the storage medium and with other hardware and software in the physical device.

[0095] Through the description of the above implementation methods, the technical personnel in this field can clearly understand that the present application can be implemented by means of software plus the necessary general hardware platform, or by hardware. By applying the technical solution of the present application, compared with the existing regular rule-based method that can only identify fraud cases with existing key features, the scheme has low timeliness, and the technical solution based on the machine learning method has a low accuracy in identifying fraud cases due to the high requirements for sample data, a social network relationship graph is constructed based on the role information and target information in the newly added claim event using the insurance business relationship; according to the social network relationship graph, a traversal starting point for constructing a closed-loop relationship graph is obtained; a plurality of closed-loop relationship graphs for a plurality of traversal starting points are constructed using the first neighbor point acquisition method based on the claim event relationship; a second neighbor point acquisition method based on the claim event relationship is used to merge the plurality of closed-loop relationship graphs to obtain a target closed-loop relationship graph, so as to obtain a technical solution for finding a fraud gang, and further find key members by calculating the shortest path. It can be seen that by regularly traversing and finding the closed-loop relationship of the association relationship between different insurance roles and insurance targets in insurance underwriting claim events under the shortest path, the key members of the target insurance fraud gang can be effectively found.

[0096] Those skilled in the art will appreciate that the accompanying drawings are only schematic diagrams of a preferred implementation scenario, and the modules or processes in the accompanying drawings are not necessarily necessary for implementing the present application. Those skilled in the art will appreciate that the modules in the system in the implementation scenario can be distributed in the system of the implementation scenario according to the implementation scenario description, or can be changed accordingly and located in one or more systems different from the present implementation scenario. The modules of the above-mentioned implementation scenario can be combined into one module, or can be further split into multiple units.

[0097] The above serial numbers of this application are only for description and do not represent the advantages and disadvantages of the implementation scenarios. The above disclosure is only a few specific implementation scenarios of this application, but this application is not limited to them, and any changes that can be thought of by technicians in this field should fall within the scope of protection of this application.

Claims

1. A method for identifying insurance fraud gangs based on social networks, characterized in that: include: Based on the role information and target information in the newly added claims events, the social network relationship diagram is constructed using the insurance business relationship; According to the social network relationship graph, obtaining a traversal starting point for constructing a closed-loop relationship graph; Using the first neighbor point acquisition method based on the claim event relationship, construct multiple closed-loop relationship graphs for multiple traversal starting points; By utilizing the second neighbor point acquisition method based on the claim event relationship, the target closed-loop relationship graph is obtained by merging the multiple closed-loop relationship graphs, and the node information in the target closed-loop relationship graph is used to characterize the target insurance fraud gang.

2. The method according to claim 1, characterized in that The social network relationship diagram includes: A first social network relationship graph is constructed by traversing the newly added claim settlement events, wherein the first social network relationship graph includes the newly added claim settlement events; or / and, By traversing the historical claims events and the newly added claims events, a second social network relationship graph is constructed, and the second social network relationship graph includes the newly added claims events and the historical claims events.

3. The method according to claim 1, characterized in that The step of constructing a social network relationship diagram using insurance business relationships based on role information and target information in the newly added claim settlement event includes: Using the role information in the newly added claim settlement event and the vehicle information in the target information as nodes in the social network relationship graph; For each node, obtain the number of claim settlement events associated with the node, and store the number of claim settlement events on the node; The nodes are connected using insurance business relationships to construct a social network relationship graph.

4. The method according to claim 3, characterized in that The target information includes the claim amount corresponding to each claim event. The step of obtaining a traversal starting point for constructing a closed-loop relationship graph according to the social network relationship graph includes: Obtaining a node corresponding to the maximum number of claim settlement events in the social network relationship graph and using it as a traversal starting point for constructing a closed-loop relationship graph; or, A plurality of nodes corresponding to the maximum number of claim settlement events in the social network relationship graph are obtained, and a traversal starting point for constructing a closed-loop relationship graph is determined according to the claim settlement amount information in the plurality of nodes.

5. The method according to claim 1, characterized in that The step of constructing multiple closed-loop relationship graphs for multiple traversal starting points by using the first neighbor point acquisition method based on the claim event relationship includes: Using the first neighbor point acquisition method based on the claim event relationship, a relationship chain based on the current traversal starting point is constructed; Acquire indirect neighboring points in the relationship chain that have the same insurance business relationship as the traversal starting point, and construct multiple closed-loop relationship graphs for the current traversal starting point based on the indirect neighboring points; The first neighbor point acquisition method is to acquire multiple nodes having the same insurance business relationship with the traversal starting point, and determine the direct neighbor points of the traversal starting point according to the number of claim settlement events corresponding to the same insurance business relationship; The direct neighboring point means that there are no other nodes between the traversal starting point and the neighboring point, and the indirect neighboring point means that there are other nodes between the traversal starting point and the neighboring point.

6. The method according to claim 1, characterized in that The step of obtaining a target closed-loop relationship graph by merging the multiple closed-loop relationship graphs using the second neighbor point acquisition method based on the claim event relationship includes: If there are associated nodes of the same claim event relationship in at least two of the multiple closed-loop relationship graphs, the at least two closed-loop relationship graphs are merged based on the associated nodes to obtain a target closed-loop relationship graph.

7. The method according to claim 1, characterized in that The method further comprises: By traversing any two nodes in the target closed-loop relationship graph, determining the shortest distance between the any two nodes, and constructing an adjacency matrix based on the shortest distance; Calculate the sum of the shortest distances of each node in the target closed-loop relationship graph according to the adjacency matrix; The nodes with the shortest distance and meeting the set conditions are regarded as the key members of the target insurance fraud gang.

8. A social network-based insurance fraud gang identification system, characterized in that: include: The first construction module is used to construct a social network relationship diagram using insurance business relationships based on role information and target information in the newly added claim event; A traversal module, used to obtain a traversal starting point for constructing a closed-loop relationship graph according to the social network relationship graph; A second construction module is used to construct multiple closed-loop relationship graphs for multiple traversal starting points by using a first neighbor point acquisition method based on a claim event relationship; A merging module is used to utilize a second neighbor point acquisition method based on the claim event relationship to merge the multiple closed-loop relationship graphs to obtain a target closed-loop relationship graph, wherein the node information in the target closed-loop relationship graph is used to characterize the target insurance fraud gang.

9. A computer storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the social network-based insurance fraud gang identification method described in any one of claims 1 to 7 is implemented.

10. A computer device comprising a storage medium, a processor, and a computer program stored in the storage medium and executable on the processor, characterized in that: When the processor executes the program, the social network-based insurance fraud gang identification method described in any one of claims 1 to 7 is implemented.

Images