Loan fraud mode identification method and system

Abstract

The invention discloses a loan fraud mode identification method and system, and relates to the technical field of fraud identification. Comprising the steps that loan information of a target user is acquired, and the loan information comprises requester information and current position information for initiating a loan request; obtaining historical position information of the requester by using an information management server according to the requester information; constructing and training a graph neural network model; constructing a position information map according to the current position information and the historical position information; inputting the position information atlas into a graph neural network model to obtain a first fraud risk factor of the loan information; and inputting the obtained first fraud risk factor of the loan information into a fraud identification model for fraud identification to obtain a fraud identification result of the loan information. The loan fraud identification level can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of fraud identification, and more specifically to a loan fraud pattern identification method and system. Background Art

[0002] With the development of economy, there are more and more frauds in the financial field. Banks and financial institutions lose tens to hundreds of billions of dollars every year due to frauds. In addition, a lot of manpower and financial resources are needed every year to detect and prevent financial frauds. Among them, loan application fraud is the most common fraud and the main factor causing losses to banks and financial institutions.

[0003] At present, fraud is common around us, and fraud in the financial industry is endless. With the continuous development of Internet finance, many financial activities have been carried out online, and financial loans are one of them. This inevitably leads to more fraud. Therefore, while ensuring that online users can quickly obtain the ultimate user experience of financial loans, it has become a top priority to prevent and control fraud risks. In addition, the existing fraud identification methods are not specifically applicable to loan behaviors. Fraud identification in loan behaviors is a technical problem that needs to be solved urgently.

[0004] Therefore, it is an urgent problem for those skilled in the art to propose a loan fraud pattern recognition method and system to solve the difficulties existing in the prior art. Summary of the invention

[0005] In view of this, the present invention provides a loan fraud pattern recognition method and system for solving the technical problems existing in the prior art.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] A loan fraud pattern recognition method comprises the following steps:

[0008] Obtain the target user's loan information, which includes the applicant information and the current location information of the person initiating the loan request;

[0009] Obtain the historical location information of the requester using the information management server according to the requester information;

[0010] Build and train graph neural network models;

[0011] Construct a location information map based on current location information and historical location information;

[0012] Input the location information graph into the graph neural network model to obtain the first fraud risk factor of the loan information;

[0013] The obtained first fraud risk factor of the loan information is input into a fraud identification model for fraud identification to obtain a fraud identification result of the loan information.

[0014] Optionally, the requester information includes an event identifier of the requested event and an identity code of the requester.

[0015] Optionally, after obtaining the loan information of the target user, it also includes: authenticating the identity code of the requester, obtaining business reference information associated with the request event and performing a business assessment based on the event identifier of the request event and the verified identity code of the requester, and confirming the fraud risk of the loan information.

[0016] Optionally, the specific content of obtaining the historical location information of the requester using the information management server according to the requester information is: the information management server stores the location information by obtaining the requester information to form a database, and obtains the historical location information of the requester by querying the database.

[0017] Optionally, the specific content of constructing the location information map based on the current location information and historical location information is:

[0018] Define nodes and edges based on current location information and historical locations, and use a graph database or graph computing framework to generate a location information graph.

[0019] Optionally, the first fraud risk factor of the obtained loan information is input into a fraud identification model for fraud identification. After obtaining the fraud identification result of the loan information, it also includes: setting a threshold to divide the fraud identification result into high-risk loans and low-risk loans; for loans less than or equal to the threshold, it is judged as a low-risk loan, judged as a safe loan and loaned; for loans greater than the threshold, it is judged as a high-risk loan, and the loan application of the target user is rejected.

[0020] A loan fraud pattern recognition system, which implements any of the above loan fraud pattern recognition methods, comprises an information acquisition module, a historical location acquisition module, a model building module, a location information map building module, a risk factor acquisition module, and an identification result module connected in sequence; wherein:

[0021] Information acquisition module: obtains the target user's loan information, which includes the applicant information and the current location information of the person initiating the loan request;

[0022] Historical location acquisition module: obtains the historical location information of the requester using the information management server according to the requester information;

[0023] Build model module: build and train graph neural network model;

[0024] Build location information map module: build location information map based on current location information and historical location information;

[0025] Risk factor acquisition module: inputs the location information map into the graph neural network model to obtain the first fraud risk factor of the loan information;

[0026] Identification result module: inputs the first fraud risk factor of the obtained loan information into the fraud identification model for fraud identification to obtain the fraud identification result of the loan information.

[0027] It can be seen from the above technical solutions that, compared with the prior art, the present invention discloses a method for identifying loan fraud patterns, which has the following beneficial effects:

[0028] 1) Through the above steps, the identification level of loan fraud can be improved;

[0029] 2) Using location information maps for model prediction can effectively detect whether someone is impersonating someone else and whether it is a temporal and spatial clustered fraud, thereby improving the level of loan fraud risk identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0031] Figure 1 A flowchart of a loan fraud pattern recognition method provided by the present invention;

[0032] Figure 2 This is a structural diagram of a loan fraud pattern recognition system provided by the present invention. DETAILED DESCRIPTION

[0033] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0034] See also Figure 1 As shown, the embodiment of the present invention discloses a loan fraud pattern identification method, comprising the following steps:

[0035] Obtain the target user's loan information, which includes the applicant information and the current location information of the person initiating the loan request;

[0036] Obtain the historical location information of the requester using the information management server according to the requester information;

[0037] Build and train graph neural network models;

[0038] Construct a location information map based on current location information and historical location information;

[0039] Input the location information graph into the graph neural network model to obtain the first fraud risk factor of the loan information;

[0040] The obtained first fraud risk factor of the loan information is input into a fraud identification model for fraud identification to obtain a fraud identification result of the loan information.

[0041] Furthermore, the requester information includes an event identifier of the requested event and an identification code of the requester.

[0042] Furthermore, after obtaining the loan information of the target user, it also includes: authenticating the identity code of the requester, obtaining business reference information associated with the request event and performing a business assessment based on the event identifier of the request event and the verified identity code of the requester, and confirming the fraud risk of the loan information.

[0043] Specifically, according to the preset quantitative conversion rules, various types of business reference information are converted into corresponding types of sub-evaluation values; and the business evaluation value of the request event is calculated based on the various sub-evaluation values.

[0044] This approach can take into account more of the various reference information associated with the request event (for example: behavior records when the request event occurs, operating environment information, network information when the event is executed, etc.), and also performs information processing for quantitative calculations, which can improve the accuracy of fraud event identification; and the fraud identification system automatically identifies fraud events when receiving a business processing request, automatically obtains relevant information and performs calculations, thereby improving the efficiency of fraud event identification.

[0045] Furthermore, the specific content of obtaining the historical location information of the requester using the information management server according to the requester information is as follows: the information management server stores the location information by obtaining the requester information to form a database, and obtains the historical location information of the requester by querying the database.

[0046] Specifically, the database is a MySQL database.

[0047] Furthermore, the specific content of constructing the location information map based on the current location information and the historical location information is as follows:

[0048] Define nodes and edges based on current location information and historical locations, and use a graph database or graph computing framework to generate a location information graph.

[0049] Specifically, the location information graph is a data representation based on a graph structure, which uses nodes and edges to show the association and evolution of location information. Node definition: Each location (including current location and historical location) is a node; Edge definition: Edges represent the relationship between locations, such as adjacent relationships, inclusion relationships, etc.

[0050] Specifically, the current location and historical locations are used to construct a location information map, which includes locations, location weights, and relationships between locations. The location weight is the number of times the requester has visited the location or preset location parameters. The location information map is input into the graph neural network model to obtain the first fraud risk factor output by the graph neural network model.

[0051] The location information graph is irregular and contains multiple unordered nodes of variable sizes. Each node in the graph has a different number of adjacent nodes. Traditional deep learning neural network models, such as neural network models for processing images, are based on the premise that each node is independent and are not suitable for graphs. For graphs, each node in the graph is related to other nodes in the graph, and this information can be used to capture the relationships between instances.

[0052] The graph neural network model is used. The graph neural network model is a neural network structure that processes graph data. It is based on the deep learning neural network to learn the topological structure and node information of the graph, and classify and identify the graph. The graph neural network inputs the location information graph into the model for identification, which can effectively detect whether someone is impersonating someone else.

[0053] Furthermore, the first fraud risk factor of the obtained loan information is input into a fraud identification model for fraud identification. After obtaining the fraud identification result of the loan information, it also includes: setting a threshold to divide the fraud identification result into high-risk loans and low-risk loans; for loans less than or equal to the threshold, it is judged as a low-risk loan, judged as a safe loan and loaned; for loans greater than the threshold, it is judged as a high-risk loan, and the loan application of the target user is rejected.

[0054] Specifically, it also includes a storage medium, the storage medium includes stored instructions, wherein when the instructions are executed, the device where the storage medium is located is controlled to perform any of the above-mentioned loan fraud pattern identification.

[0055] An electronic device includes a memory and one or more instructions, wherein the one or more instructions are stored in the memory and configured to be executed by one or more processors to perform any of the above-mentioned loan fraud pattern identification.

[0056] Specifically, the memory includes at least one type of readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, etc.

[0057] In some embodiments, the memory may be an internal storage unit of a computer device; in other embodiments, the memory may also be an external storage device of the computer device. Of course, the memory may also include both an internal storage unit of a computer device and an external storage device thereof. The memory may also be used to temporarily store various types of data that have been output or are to be output.

[0058] and Figure 1 Corresponding to the method described above, the present invention also provides a loan fraud pattern recognition system for Figure 1 The specific implementation of the method in the above example is shown in the schematic diagram of its structure. Figure 2 As shown, it includes an information acquisition module, a historical location acquisition module, a model building module, a location information map building module, a risk factor acquisition module, and an identification result module connected in sequence; wherein,

[0059] Information acquisition module: obtains the target user's loan information, which includes the applicant information and the current location information of the person initiating the loan request;

[0060] Historical location acquisition module: obtains the historical location information of the requester using the information management server according to the requester information;

[0061] Build model module: build and train graph neural network model;

[0062] Build location information map module: build location information map based on current location information and historical location information;

[0063] Risk factor acquisition module: inputs the location information map into the graph neural network model to obtain the first fraud risk factor of the loan information;

[0064] Identification result module: inputs the first fraud risk factor of the obtained loan information into the fraud identification model for fraud identification to obtain the fraud identification result of the loan information.

[0065] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0066] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A loan fraud pattern recognition method, characterized in that: The following steps are involved: Obtain the target user's loan information, which includes the applicant information and the current location information of the person initiating the loan request; Obtain the historical location information of the requester using the information management server according to the requester information; Build and train graph neural network models; Construct a location information map based on current location information and historical location information; Input the location information graph into the graph neural network model to obtain the first fraud risk factor of the loan information; The obtained first fraud risk factor of the loan information is input into a fraud identification model for fraud identification to obtain a fraud identification result of the loan information.

2. A loan fraud pattern recognition method according to claim 1, characterized in that: The requester information includes the event identifier of the requested event and the identity code of the requester.

3. A loan fraud pattern recognition method according to claim 2, characterized in that: After obtaining the target user's loan information, it also includes: authenticating the identity code of the requester, obtaining the business reference information associated with the request event and performing a business assessment based on the event identifier of the request event and the verified identity code of the requester, and confirming the fraud risk of the loan information.

4. A loan fraud pattern recognition method according to claim 1, characterized in that: The specific content of obtaining the historical location information of the requester using the information management server according to the requester information is as follows: the information management server stores the location information by obtaining the requester information to form a database, and obtains the historical location information of the requester by querying the database.

5. A loan fraud pattern recognition method according to claim 1, characterized in that: The specific contents of constructing the location information map based on the current location information and historical location information are as follows: Define nodes and edges based on current location information and historical locations, and use a graph database or graph computing framework to generate a location information graph.

6. A loan fraud pattern recognition method according to claim 1, characterized in that: The first fraud risk factor of the obtained loan information is input into a fraud identification model for fraud identification. After the fraud identification result of the loan information is obtained, the method also includes: setting a threshold to divide the fraud identification result into high-risk loans and low-risk loans; for loans less than or equal to the threshold, it is judged as a low-risk loan, judged as a safe loan and loaned; for loans greater than the threshold, it is judged as a high-risk loan, and the loan application of the target user is rejected.

7. A loan fraud pattern recognition system, characterized in that A loan fraud pattern recognition method according to any one of claims 1 to 6 is applied, comprising an information acquisition module, a historical location acquisition module, a model building module, a location information map building module, a risk factor acquisition module, and an identification result module connected in sequence; wherein, Information acquisition module: obtains the target user's loan information, which includes the applicant information and the current location information of the person initiating the loan request; Historical location acquisition module: obtains the historical location information of the requester using the information management server according to the requester information; Build model module: build and train graph neural network model; Build location information map module: build location information map based on current location information and historical location information; Risk factor acquisition module: inputs the location information map into the graph neural network model to obtain the first fraud risk factor of the loan information; Identification result module: inputs the first fraud risk factor of the obtained loan information into the fraud identification model for fraud identification to obtain the fraud identification result of the loan information.

Images