Business risk management and control method, device and equipment and storage medium thereof

Abstract

The embodiment of the invention belongs to the technical field of business safety supervision, is applied to a financial business risk management and control scene, and relates to a business risk management and control method, device and equipment and a storage medium thereof. Extracting business coming and going features and business attribute features; and inputting the business coming and going characteristics and the business attribute characteristics into a business risk representation model to obtain a business relation graph and risk scores and risk types of all account nodes in the business relation graph, applying the business risk management and control method to a financial or bank transaction scene, identifying transaction risks, and improving the business risk management and control efficiency. The economic loss of customers is avoided; meanwhile, by adopting the business risk management and control method, risk accounts can be screened out from unspecific accounts, management and control processing can be performed on the risk accounts in time, a financial supervision department can be assisted to screen financial crime cases, the method is more intelligent and automatic, an artificial intelligence machine learning mode is combined, and the risk identification efficiency and accuracy are improved.

Description

Technical Field

[0001] The present application relates to the field of business security supervision technology, and is applied to financial business risk management scenarios, and in particular to a business risk management method, device, equipment and storage medium thereof. Background Art

[0002] With the deepening development of mobile Internet technology and the continuous upgrading and updating of payment models and payment scenarios, telecommunications network fraud has shown characteristics such as being new, networked, and concentrated.

[0003] In order to protect the property safety of the people, the police and banks need to work together. Screening and risk control in massive amounts of customer information and transaction information consumes a lot of manpower and material resources. Although, at present, in order to reduce the consumption of manpower and material resources, some risk identification methods have emerged, but these risk identification methods are mostly based on module community discovery algorithms (Louva in algorithm, Leiden algorithm, etc.), with fixed indicators as the evaluation criteria. The subsequent maintenance and upgrading of the model mainly relies on expert knowledge. If new methods emerge, the recognition efficiency of these risk identification methods is low, and they cannot be combined with features for deep mining, resulting in low risk identification efficiency and accuracy. Summary of the invention

[0004] The purpose of the embodiments of the present application is to propose a business risk management method, device, equipment and storage medium thereof to solve the problem that the existing risk identification methods have low recognition efficiency and cannot be combined with features for deep mining, resulting in low risk identification efficiency and accuracy.

[0005] In order to solve the above technical problems, the embodiment of the present application provides a business risk management method, which adopts the following technical solutions:

[0006] A business risk management method comprises the following steps:

[0007] Acquire the data to be detected, wherein the data to be detected includes business relationship data of the batch accounts to be detected;

[0008] Performing feature extraction on the data to be detected, extracting business transaction features and business attribute features, wherein the business transaction features include business flow information between the batch accounts to be detected, and the business attribute features include business account information, business amount information, business frequency information, and business time information of the batch accounts to be detected;

[0009] Input the business transaction characteristics and the business attribute characteristics into a pre-trained business risk characterization model to obtain a business relationship graph output by the business risk characterization model, as well as risk scores and risk types of all account nodes in the business relationship graph;

[0010] According to the risk scores and risk types of all account nodes in the business relationship graph, target account nodes whose risk scores exceed a preset risk control value and whose risk types are target control types are screened out from the business relationship graph;

[0011] Send the business account information corresponding to all target account nodes to the preset risk control management terminal.

[0012] Furthermore, the business risk characterization model includes a risk feature identification component, a business relationship graph construction component and a risk type identification component. Before executing the step of inputting the business transaction features and the business attribute features into the pre-trained business risk characterization model to obtain the business relationship graph output by the business risk characterization model, and the risk scores and risk types of all account nodes in the business relationship graph, the method further includes:

[0013] Acquire a training data set, wherein the training data set includes positive and negative sample data, the positive sample data includes business relationship data of a batch of labeled risky accounts, and the negative sample data includes business relationship data of a batch of labeled non-risky accounts;

[0014] Performing feature extraction on the training data set to obtain business attribute features contained in all positive and negative sample data and business transaction features contained in all positive and negative sample data;

[0015] Inputting the business attribute features contained in all the positive and negative sample data into the risk feature identification component to be trained for learning and training, and learning to obtain a trained risk feature identification component;

[0016] Inputting the business attribute features contained in all the positive and negative sample data and the business transaction features contained in all the positive and negative sample data into the business relationship graph construction component, and obtaining the business transaction relationship graph constructed by the business relationship graph construction component according to the training data set;

[0017] The business attribute features and business transaction relationships contained in all the positive and negative sample data Figure 1 The data are input into the risk type identification component to be trained, and learning training is performed on the risk type identification component to be trained to obtain a trained risk type identification component.

[0018] Furthermore, the step of inputting the business attribute features contained in all the positive and negative sample data into the risk feature identification component to be trained for learning and training, and learning to obtain the trained risk feature identification component specifically includes:

[0019] According to the positive and negative annotation knowledge of samples, the business attribute features contained in all positive and negative sample data are sorted out, and the business attribute features corresponding to the positive sample data and the business attribute features corresponding to the negative sample data are sorted out respectively;

[0020] By comparing the business attribute features corresponding to the positive sample data with the business attribute features corresponding to the negative sample data, the detection features for risk identification are learned;

[0021] According to the feature data mapping of all detection features in the positive sample data, all detection features are sorted by importance, and the most important detection features with a preset ratio are selected;

[0022] Setting risk feature detection nodes for the risk feature identification component based on the most important detection features of the preset ratio, and setting risk detection weights for all risk feature detection nodes according to the importance ranking and a preset weight allocation strategy;

[0023] The risk feature identification component in which the risk feature detection nodes and the risk detection weights have been set is used as the trained risk feature identification component.

[0024] Furthermore, the step of ranking all detection features by importance according to feature data mapping of all detection features in positive sample data, and screening out the most important detection features of a preset proportion, specifically includes:

[0025] Count the number of samples corresponding to all detection features in the positive sample data;

[0026] According to the relationship of sample quantity from high to low, all the detection features are sorted in descending order to obtain the importance sorting result;

[0027] According to a preset ratio value, extracting detection features of a target ratio from the importance sorting result as the most important detection features of the preset ratio;

[0028] The step of setting a risk feature detection node for the risk feature identification component based on the most important detection feature of the preset ratio, and setting a risk detection weight for all risk feature detection nodes according to the importance ranking and a preset weight allocation strategy, specifically includes:

[0029] Setting the most important detection features of the preset proportion as risk feature detection nodes of the risk feature identification component;

[0030] According to the sequence information of the most important detection features of the preset proportion in the importance ranking result, risk detection weights are set for all risk feature detection nodes, wherein the preset weight allocation strategy includes that the higher the sequence information of the detection feature in the importance ranking result, the greater the risk detection weight corresponding to the detection feature.

[0031] Furthermore, the step of inputting the business attribute features contained in all the positive and negative sample data and the business transaction features contained in all the positive and negative sample data into the business relationship graph construction component to obtain the business transaction relationship graph constructed by the business relationship graph construction component according to the training data set specifically includes:

[0032] Obtaining account information corresponding to all positive and negative sample data respectively from the business attribute features contained in all the positive and negative sample data;

[0033] The account information corresponding to all the positive and negative sample data is used as a graph node of the business transaction relationship graph;

[0034] Constructing edge relationships between graph nodes in the business transaction relationship graph according to the business transaction features contained in all the positive and negative sample data;

[0035] The business transaction relationship graph is generated based on the graph nodes of the business transaction relationship graph and the edge relationships between the graph nodes in the business transaction relationship graph.

[0036] Furthermore, the business attribute features and the business relationship contained in all the positive and negative sample data are Figure 1 The steps of inputting the risk type identification component to be trained, performing learning training on the risk type identification component to be trained, and obtaining the trained risk type identification component specifically include:

[0037] Based on the risk type annotation knowledge, identifying the non-group risk business accounts and the group risk business accounts in the business transaction relationship diagram;

[0038] Summarize and organize the business attribute characteristics of non-group risk business accounts to obtain summary results of the business attribute characteristics of non-group risk business accounts;

[0039] Summarize and organize the business attribute characteristics of the group-type risk business accounts to obtain the summary results of the business attribute characteristics of the group-type risk business accounts;

[0040] According to the summary result of the business attribute characteristics of the non-group risk business account, the type identification characteristics of the non-group risk business account are determined as the first type identification characteristics;

[0041] According to the summary results of the business attribute characteristics of the group-type risk business account, the type identification characteristics of the group-type risk business account are determined as the second type identification characteristics;

[0042] According to different risk types, respectively set type recognition weights for the first type recognition feature and the second type recognition feature, and set type feature recognition nodes for the risk type recognition component to be trained according to the first type recognition feature and the second type recognition feature;

[0043] The risk type identification component in which the type feature identification node and the type identification weight are set is used as the trained risk type identification component.

[0044] Furthermore, the step of inputting the business transaction characteristics and the business attribute characteristics into a pre-trained business risk characterization model to obtain a business relationship graph output by the business risk characterization model, as well as the risk scores and risk types of all account nodes in the business relationship graph, specifically includes:

[0045] The risk feature identification component performs risk feature detection on the business attribute features of all accounts to be detected, and identifies the risk feature detection nodes hit by all accounts to be detected;

[0046] The risk feature detection nodes hit by the same account to be detected are aggregated and sorted to obtain the detection node sets corresponding to all accounts to be detected;

[0047] According to the detection node sets corresponding to all the accounts to be detected and the risk detection weights corresponding to the risk feature detection nodes, the risk detection weights corresponding to the elements in the same detection node set are weighted and summed to obtain the risk detection scores corresponding to all the accounts to be detected;

[0048] The risk detection scores corresponding to all the accounts to be detected are used as the risk scores corresponding to all the accounts to be detected;

[0049] Obtaining business account information of all accounts to be detected from the business attribute characteristics;

[0050] Using the business account information of all the accounts to be detected as graph nodes of the business relationship graph;

[0051] Identify the edge relationships between all graph nodes based on the business transaction characteristics;

[0052] Generate the business relationship graph based on the graph node and the edge relationships between all graph nodes;

[0053] The risk type identification component identifies the risk type of the business attribute characteristics of all accounts to be detected, and identifies the type feature identification nodes hit by all accounts to be detected;

[0054] The type feature identification nodes hit by the same account to be detected are collected and sorted to obtain the type node sets corresponding to all the accounts to be detected;

[0055] According to the type node sets corresponding to all the accounts to be detected and the type identification weights corresponding to the type feature identification nodes, the type identification weights corresponding to the elements in the same type node set are weighted summed to obtain the type identification scores corresponding to all the accounts to be detected;

[0056] Determine the risk types corresponding to all accounts to be detected according to the type identification score and the preset type identification rules;

[0057] According to the business account information of all the accounts to be detected, the risk scores and risk types corresponding to all the accounts to be detected are marked at the corresponding graph node positions in the business relationship graph to obtain a marked business relationship graph;

[0058] The annotated business relationship diagram is outputted through the business risk characterization model.

[0059] In order to solve the above technical problems, the embodiment of the present application also provides a business risk management and control device, which adopts the following technical solution:

[0060] A business risk management and control device, comprising:

[0061] A module for acquiring data to be detected, used to acquire data to be detected, wherein the data to be detected includes business relationship data of batch accounts to be detected;

[0062] A feature extraction module is used to extract features from the data to be detected, and extract business transaction features and business attribute features, wherein the business transaction features include business flow information between the batch accounts to be detected, and the business attribute features include business account information, business amount information, business frequency information, and business time information of the batch accounts to be detected;

[0063] A model processing module, used to input the business transaction characteristics and the business attribute characteristics into a pre-trained business risk characterization model, obtain a business relationship graph output by the business risk characterization model, and risk scores and risk types of all account nodes in the business relationship graph;

[0064] A target account node screening module, configured to screen target account nodes whose risk scores exceed a preset risk control value and whose risk types are target control types from the business relationship graph according to the risk scores and risk types of all account nodes in the business relationship graph;

[0065] The business account information sending module is used to send the business account information corresponding to all target account nodes to the preset risk control management terminal.

[0066] In order to solve the above technical problems, the embodiment of the present application further provides a computer device, which adopts the following technical solution:

[0067] A computer device includes a memory and a processor, wherein the memory stores computer-readable instructions, and the processor implements the steps of the business risk management method described above when executing the computer-readable instructions.

[0068] In order to solve the above technical problems, the embodiment of the present application further provides a computer-readable storage medium, which adopts the following technical solution:

[0069] A computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the business risk management method as described above.

[0070] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0071] The business risk management method described in the embodiment of the present application obtains the data to be detected; extracts features from the data to be detected, extracts business transaction features and business attribute features; inputs the business transaction features and business attribute features into the pre-trained business risk characterization model, obtains the business relationship diagram output by the business risk characterization model and the risk scores and risk types of all account nodes in the business relationship diagram; based on the risk scores and risk types of all account nodes in the business relationship diagram, selects target account nodes whose risk scores exceed the preset risk control value and whose risk types are the target control type from the business relationship diagram; and sends the business account information corresponding to all target account nodes to the preset risk control management terminal. The business risk management method is applied to financial or banking transaction scenarios, which can identify transaction risks and avoid economic losses of customers; at the same time, the business risk management method can also screen out risk accounts in unspecified accounts, and timely manage and handle risk accounts, assisting financial regulatory authorities in screening financial crime cases, which is more intelligent and automated, and combines artificial intelligence machine learning methods to improve risk identification efficiency and accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] In order to more clearly illustrate the scheme in the present application, a brief introduction is given below to the drawings required for use in the description of the embodiments of the present application. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0073] Figure 1 is an exemplary system architecture diagram to which the present application may be applied;

[0074] Figure 2 is a flow chart of an embodiment of a business risk management method according to the present application;

[0075] Figure 3 yes Figure 2 A flowchart of a specific embodiment of performing learning and training on the business risk characterization model before step 203 is shown;

[0076] Figure 4 yes Figure 3 A flowchart of a specific embodiment of step 303 is shown;

[0077] Figure 5 yes Figure 4 A flowchart of a specific embodiment of step 403 is shown;

[0078] Figure 6 yes Figure 4 A flowchart of a specific embodiment of step 404 is shown;

[0079] Figure 7 yes Figure 3 A flowchart of a specific embodiment of step 304 is shown;

[0080] Figure 8 yes Figure 3 A flowchart of a specific embodiment of step 305 is shown;

[0081] Fig. 9 is a structural diagram of an embodiment of a business risk management and control device according to the present application;

[0082] Fig.10 It is a structural diagram of an embodiment of a computer device according to the present application. DETAILED DESCRIPTION

[0083] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by technicians in the technical field of the present application; the terms used in the specification of the application herein are only for the purpose of describing specific embodiments and are not intended to limit the present application; the terms "including" and "having" and any variations thereof in the specification and claims of the present application and the above-mentioned drawings are intended to cover non-exclusive inclusions. The terms "first", "second", etc. in the specification and claims of the present application or the above-mentioned drawings are used to distinguish different objects, not to describe a specific order.

[0084] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0085] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings.

[0086] like Figure 1 As shown, the system architecture 100 may include a terminal device 101, a network 102 and a server 103. The terminal device 101 may be a laptop 1011, a tablet computer 1012 or a mobile phone 1013. The network 102 is used to provide a medium for a communication link between the terminal device 101 and the server 103. The network 102 may include various connection types, such as wired, wireless communication links or optical fiber cables.

[0087] The user can use the terminal device 101 to interact with the server 103 through the network 102 to receive or send messages, etc. Various communication client applications can be installed on the terminal device 101, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc.

[0088] The terminal device 101 can be various electronic devices with a display screen and supporting web browsing. In addition to the laptop computer 1011, the tablet computer 1012 or the mobile phone 1013, the terminal device 101 can also be an e-book reader, an MP3 player (Moving Picture Experts Group Audio Layer III), an MP4 (Moving Picture Experts Group Audio Layer IV) player, a laptop computer and a desktop computer, etc.

[0089] The server 103 may be a server that provides various services, such as a background server that provides support for a web page displayed on the terminal device 101 .

[0090] It should be noted that the business risk management method provided in the embodiment of the present application is generally executed by a terminal device, and accordingly, the business risk management device is generally provided in the terminal device.

[0091] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to the implementation requirements.

[0092] Continue to refer Figure 2 , shows a flow chart of an embodiment of a business risk management method according to the present application. The business risk management method comprises the following steps:

[0093] Step 201 : obtaining data to be detected, wherein the data to be detected includes business relationship data of batch accounts to be detected.

[0094] In this embodiment, the batch accounts to be detected, for example, the batch bank transaction accounts to be detected. If the batch accounts to be detected are the batch bank transaction accounts to be detected, then the business relationship data includes transfer record information, remittance information, etc., and also includes account information, amount information, transfer time information, transfer number information, etc. with transaction relationship. The transaction account can be a personal account or a company account, which is not limited here.

[0095] Applying the business risk management method to financial or banking transaction scenarios can identify transaction risks and avoid financial losses for customers. At the same time, the business risk management method can also screen out risky accounts from unspecified accounts, manage and handle risky accounts in a timely manner, and assist financial regulatory authorities in screening financial crime cases.

[0096] Step 202, feature extraction is performed on the data to be detected, and business transaction features and business attribute features are extracted. The business transaction features include business flow information between the batch accounts to be detected, and the business attribute features include business account information, business amount information, business frequency information and business time information of the batch accounts to be detected.

[0097] By performing feature extraction on the data to be detected, business transaction features and business attribute features are extracted, so that the business transaction features and business attribute features can be processed by a business risk characterization model later, making data processing more intelligent.

[0098] Step 203: input the business transaction characteristics and the business attribute characteristics into the pre-trained business risk characterization model to obtain the business relationship graph output by the business risk characterization model, as well as the risk scores and risk types of all account nodes in the business relationship graph.

[0099] By directly inputting the business transaction characteristics and the business attribute characteristics into the pre-trained business risk characterization model, the business relationship graph output by the business risk characterization model, as well as the risk scores and risk types of all account nodes in the business relationship graph are obtained, and artificial intelligence is used for processing, thereby improving processing efficiency and making it more intelligent and automated.

[0100] Step 204: based on the risk scores and risk types of all account nodes in the business relationship graph, target account nodes whose risk scores exceed a preset risk control value and whose risk types are target control types are screened out from the business relationship graph.

[0101] In this embodiment, the preset risk control value can be understood as a preset risk score threshold, and the target control type can also be understood as a pre-set risk type, such as group type and non-group type. Generally speaking, taking telecommunications fraud as an example, it often involves gang-type and group-type capital turnover transactions, which may be interspersed with multiple transfers and money laundering operations. The transfer account information used here is very easy to be identified as a group-type risk business account due to the intensive and frequent transaction correlations between each other; and a one-time illegal fundraising case, for example: multiple accounts transfer funds to the same account in almost the same time period, the same account here is very easy to be identified as a non-group-type risk business account. By screening out the target account nodes whose risk scores exceed the preset risk control value and whose risk type is the target control type from the business relationship diagram, the risk control management end can perform risk control management.

[0102] Step 205: Send the business account information corresponding to all target account nodes to the preset risk control management terminal.

[0103] In this embodiment, by acquiring the data to be detected; extracting features from the data to be detected, extracting business transaction features and business attribute features; inputting the business transaction features and business attribute features into the pre-trained business risk characterization model, obtaining the business relationship graph output by the business risk characterization model and the risk scores and risk types of all account nodes in the business relationship graph; based on the risk scores and risk types of all account nodes in the business relationship graph, screening out target account nodes whose risk scores exceed the preset risk control value and whose risk type is the target control type from the business relationship graph; sending the business account information corresponding to all target account nodes to the preset risk control management terminal. The business risk control method is applied to financial or banking transaction scenarios, which can identify transaction risks and avoid economic losses of customers; at the same time, the business risk control method can also screen out risk accounts in unspecified accounts, and timely control and handle risk accounts, assisting financial regulatory authorities in screening financial crime cases.

[0104] In this embodiment, the business risk characterization model includes a risk feature identification component, a business relationship graph construction component, and a risk type identification component. Specifically, the business risk characterization model can be learned and trained by a classification prediction model based on a graph neural network. The use of a graph neural network can not only realize the identification function of the risk feature identification component and the risk type identification component, but also realize the relationship graph construction function of the business relationship graph construction component.

[0105] Continue to refer Figure 3 , Figure 3 yes Figure 2 The flowchart of a specific embodiment of learning and training the business risk characterization model before step 203 is shown. In some optional implementations, learning and training the business risk characterization model before step 203 includes the following steps:

[0106] Step 301, obtaining a training data set, wherein the training data set includes positive and negative sample data, the positive sample data includes business relationship data of a batch of labeled risky accounts, and the negative sample data includes business relationship data of a batch of labeled non-risky accounts;

[0107] Step 302, extracting features from the training data set to obtain business attribute features contained in all positive and negative sample data and business transaction features contained in all positive and negative sample data;

[0108] Step 303: input the business attribute features contained in all the positive and negative sample data into the risk feature identification component to be trained for learning and training, and obtain a trained risk feature identification component;

[0109] Step 304: input the business attribute features contained in all the positive and negative sample data and the business transaction features contained in all the positive and negative sample data into the business relationship graph construction component to obtain a business transaction relationship graph constructed by the business relationship graph construction component according to the training data set;

[0110] Step 305: The business attribute features and the business relationship contained in all the positive and negative sample data are Figure 1 The data are input into the risk type identification component to be trained, and learning training is performed on the risk type identification component to be trained to obtain a trained risk type identification component.

[0111] In this embodiment, the business risk characterization model is learned and trained through steps 301 to 305. Specifically, the recognition functions of the risk feature identification component and the risk type identification component in the business risk characterization model are learned and trained. At the same time, the business relationship graph construction component in the business risk characterization model can also be learned and trained, thereby obtaining a pre-trained business risk characterization model for use in subsequent actual business risk identification.

[0112] Continue to refer Figure 4 , Figure 4 yes Figure 3 The flowchart of a specific embodiment of step 303 is shown. In some optional implementations, step 303 includes the following steps:

[0113] Step 401, sorting out the business attribute features contained in all positive and negative sample data according to the sample positive and negative labeling knowledge, and sorting out the business attribute features corresponding to the positive sample data and the business attribute features corresponding to the negative sample data respectively;

[0114] Step 402, by comparing the business attribute features corresponding to the positive sample data with the business attribute features corresponding to the negative sample data, the detection features for risk identification are learned;

[0115] Step 403, sorting all detection features by importance according to the feature data mapping of all detection features in the positive sample data, and screening out the most important detection features of a preset proportion;

[0116] Step 404, setting a risk feature detection node for the risk feature identification component based on the most important detection features of the preset ratio, and setting a risk detection weight for all risk feature detection nodes according to the importance ranking and a preset weight allocation strategy;

[0117] Step 405: The risk feature identification component in which the risk feature detection nodes and the risk detection weights are set is used as the trained risk feature identification component.

[0118] Specifically, when learning and training the risk feature identification component, its main learning purpose is to learn the risk detection features, and to learn the risk detection weights corresponding to the risk detection features based on the positive and negative annotation knowledge of the samples, so as to facilitate risk detection of unspecific detection data, that is, detection data of unknown risk or not, in subsequent actual use.

[0119] Continue to refer Figure 5 , Figure 5 yes Figure 4 The flowchart of a specific embodiment of step 403 is shown. In some optional implementations, step 403 includes the following steps:

[0120] Step 501, counting the number of samples corresponding to all detection features in the positive sample data;

[0121] Step 502, sorting all the detection features in descending order according to the number of samples from high to low, to obtain an importance sorting result;

[0122] Step 503: extracting detection features of a target ratio from the importance sorting result according to a preset ratio value as the most important detection features of the preset ratio.

[0123] Continue to refer Figure 6 , Figure 6 yes Figure 4 The flowchart of a specific embodiment of step 404 is shown. In some optional implementations, step 404 includes the following steps:

[0124] Step 601, setting the most important detection feature of the preset proportion as the risk feature detection node of the risk feature identification component;

[0125] Step 602, setting risk detection weights for all risk feature detection nodes according to the sequence information of the most important detection features of the preset proportion in the importance ranking result, wherein the preset weight allocation strategy includes that the higher the sequence information of the detection feature in the importance ranking result, the greater the risk detection weight corresponding to the detection feature.

[0126] Specifically, Figure 5 and Figure 6 The steps shown are actually to optimize the risk detection features learned by the risk feature identification component, delete non-important features, and retain only important features, thereby improving the risk detection efficiency of the model.

[0127] Continue to refer Figure 7 , Figure 7 yes Figure 3The flowchart of a specific embodiment of step 304 is shown. In some optional implementations, step 304 includes the following steps:

[0128] Step 701, acquiring the account information corresponding to all the positive and negative sample data respectively from the business attribute features contained in all the positive and negative sample data;

[0129] Step 702, using the account information corresponding to all the positive and negative sample data as graph nodes of the business transaction relationship graph;

[0130] Step 703, constructing edge relationships between graph nodes in the business transaction relationship graph according to the business transaction features contained in all the positive and negative sample data;

[0131] Step 704: Generate the business transaction relationship graph based on the graph nodes of the business transaction relationship graph and the edge relationships between the graph nodes in the business transaction relationship graph.

[0132] Specifically, by generating the business transaction relationship graph, it is easy to identify risk types from the dimension of business transactions. It can be understood that if multiple graph nodes, that is, multiple account information, frequently conduct business transactions, and the multiple accounts are accounts with high risk scores, the multiple account information can be divided into group-type risk business accounts within the same group. It is convenient to combine graph knowledge to identify and predict business risks, and improve the interpretability of the model.

[0133] Continue to refer Figure 8 , Figure 8 yes Figure 3 The flowchart of a specific embodiment of step 305 is shown. In some optional implementations, step 305 includes the following steps:

[0134] Step 801, based on risk type annotation knowledge, identifying non-group risk business accounts and group risk business accounts in the business transaction relationship diagram;

[0135] Step 802, summarizing and arranging the business attribute characteristics of the non-group risk business accounts to obtain a summary result of the business attribute characteristics of the non-group risk business accounts;

[0136] Step 803, summarizing and arranging the business attribute characteristics of the group-type risk business accounts to obtain a summary result of the business attribute characteristics of the group-type risk business accounts;

[0137] Step 804, according to the summary result of the business attribute characteristics of the non-group risk business account, determine the type identification characteristics of the non-group risk business account as the first type identification characteristics;

[0138] Step 805, according to the summary result of the business attribute characteristics of the group-type risk business account, determine the type identification characteristics of the group-type risk business account as the second type identification characteristics;

[0139] Step 806: setting type recognition weights for the first type recognition feature and the second type recognition feature respectively according to different risk types, and setting type feature recognition nodes for the risk type recognition component to be trained according to the first type recognition feature and the second type recognition feature;

[0140] Step 807: The risk type identification component in which the type feature identification node and the type identification weight are set is used as the trained risk type identification component.

[0141] Specifically, by summarizing the business attribute characteristics of non-group risk business accounts and the business attribute characteristics of group risk business accounts respectively, the type identification characteristics of non-group risk business accounts and the type identification characteristics of group risk business accounts can be obtained from a group perspective, so that when actually identifying the risk type, it can be directly identified based on the type identification characteristics.

[0142] In this embodiment, based on the business transaction relationship graph, higher-dimensional features can also be generated according to the business transaction characteristics between graph nodes, such as: the number of business transactions between graph nodes, and the position characteristics of graph nodes in the business transaction relationship graph. The position characteristics can be understood as edge nodes and non-edge nodes, thereby combining higher-dimensional features to identify risk types.

[0143] In some optional implementations, step 203, inputting the business transaction characteristics and the business attribute characteristics into a pre-trained business risk characterization model, obtaining a business relationship graph output by the business risk characterization model, and risk scores and risk types of all account nodes in the business relationship graph, includes the following steps:

[0144] The risk feature identification component performs risk feature detection on the business attribute features of all accounts to be detected, and identifies the risk feature detection nodes hit by all accounts to be detected;

[0145] The risk feature detection nodes hit by the same account to be detected are aggregated and sorted to obtain the detection node sets corresponding to all accounts to be detected;

[0146] According to the detection node sets corresponding to all the accounts to be detected and the risk detection weights corresponding to the risk feature detection nodes, the risk detection weights corresponding to the elements in the same detection node set are weighted and summed to obtain the risk detection scores corresponding to all the accounts to be detected;

[0147] The risk detection scores corresponding to all the accounts to be detected are used as the risk scores corresponding to all the accounts to be detected;

[0148] Obtaining business account information of all accounts to be detected from the business attribute characteristics;

[0149] Using the business account information of all the accounts to be detected as graph nodes of the business relationship graph;

[0150] Identify the edge relationships between all graph nodes based on the business transaction characteristics;

[0151] Generate the business relationship graph based on the graph node and the edge relationships between all graph nodes;

[0152] The risk type identification component identifies the risk type of the business attribute characteristics of all accounts to be detected, and identifies the type feature identification nodes hit by all accounts to be detected;

[0153] The type feature identification nodes hit by the same account to be detected are collected and sorted to obtain the type node sets corresponding to all the accounts to be detected;

[0154] According to the type node sets corresponding to all the accounts to be detected and the type identification weights corresponding to the type feature identification nodes, the type identification weights corresponding to the elements in the same type node set are weighted summed to obtain the type identification scores corresponding to all the accounts to be detected;

[0155] According to the type identification score and the preset type identification rule, the risk types corresponding to all the accounts to be detected are determined, wherein the preset type identification rule, for example: if the type identification score of the first type identification feature corresponding to the current account to be detected is greater than the type identification score of the first type identification feature, and the risk score of the current account to be detected is greater than the preset risk grouping threshold, then the current account to be detected is a non-group risk business account; conversely, if the type identification score of the first type identification feature corresponding to the current account to be detected is less than the type identification score of the first type identification feature, and the risk score of the current account to be detected is greater than the preset risk grouping threshold, then the current account to be detected is a group risk business account;

[0156] According to the business account information of all the accounts to be detected, the risk scores and risk types corresponding to all the accounts to be detected are marked at the corresponding graph node positions in the business relationship graph to obtain a marked business relationship graph;

[0157] The annotated business relationship diagram is outputted through the business risk characterization model.

[0158] This application obtains the data to be tested; extracts features from the data to be tested, extracts business transaction features and business attribute features; inputs the business transaction features and business attribute features into the pre-trained business risk characterization model, obtains the business relationship diagram output by the business risk characterization model and the risk scores and risk types of all account nodes in the business relationship diagram; based on the risk scores and risk types of all account nodes in the business relationship diagram, selects target account nodes whose risk scores exceed the preset risk control value and whose risk type is the target control type from the business relationship diagram; and sends the business account information corresponding to all target account nodes to the preset risk control management terminal. The business risk control method is applied to financial or banking transaction scenarios, which can identify transaction risks and avoid economic losses of customers; at the same time, the business risk control method can also screen out risk accounts in unspecified accounts, and timely control and handle risk accounts, assisting financial regulatory authorities in screening financial crime cases, which is more intelligent and automated, combined with artificial intelligence machine learning methods, and improves risk identification efficiency and accuracy.

[0159] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Artificial Intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0160] AI basic technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. AI software technologies mainly include computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0161] In the embodiment of the present application, by acquiring the data to be detected; extracting features from the data to be detected, extracting business transaction features and business attribute features; inputting the business transaction features and business attribute features into the pre-trained business risk characterization model, obtaining the business relationship diagram output by the business risk characterization model and the risk scores and risk types of all account nodes in the business relationship diagram; based on the risk scores and risk types of all account nodes in the business relationship diagram, screening out the target account nodes whose risk scores exceed the preset risk control value and whose risk type is the target control type from the business relationship diagram; sending the business account information corresponding to all target account nodes to the preset risk control management terminal. The business risk control method is applied to financial or banking transaction scenarios, which can identify transaction risks and avoid economic losses of customers; at the same time, the business risk control method can also screen out risk accounts in unspecified accounts, and timely control and handle risk accounts, assisting financial regulatory authorities in screening financial crime cases, which is more intelligent and automated, combined with artificial intelligence machine learning methods, and improves risk identification efficiency and accuracy.

[0162] Further references Fig. 9 , as a response to the above Figure 2 In order to realize the method shown in the figure, the present application provides an embodiment of a business risk management and control device, and the device embodiment is Figure 2 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.

[0163] like Fig. 9 As shown, the business risk management and control device 9 described in this embodiment includes: a to-be-detected data acquisition module 901, a feature extraction module 902, a model processing module 903, a target account node screening module 904 and a business account information sending module 905. Among them:

[0164] The to-be-detected data acquisition module 901 is used to acquire the to-be-detected data, wherein the to-be-detected data includes the business relationship data of the to-be-detected batch accounts;

[0165] A feature extraction module 902 is used to extract features from the data to be detected, and extract business transaction features and business attribute features, wherein the business transaction features include business flow information between the batch accounts to be detected, and the business attribute features include business account information, business amount information, business frequency information, and business time information of the batch accounts to be detected;

[0166] The model processing module 903 is used to input the business transaction characteristics and the business attribute characteristics into the pre-trained business risk characterization model to obtain the business relationship graph output by the business risk characterization model, as well as the risk scores and risk types of all account nodes in the business relationship graph;

[0167] The target account node screening module 904 is used to screen out target account nodes whose risk scores exceed a preset risk control value and whose risk types are target control types from the business relationship graph according to the risk scores and risk types of all account nodes in the business relationship graph;

[0168] The business account information sending module 905 is used to send the business account information corresponding to all target account nodes to the preset risk control management terminal.

[0169] This application obtains the data to be tested; extracts features from the data to be tested, extracts business transaction features and business attribute features; inputs the business transaction features and business attribute features into the pre-trained business risk characterization model, obtains the business relationship diagram output by the business risk characterization model and the risk scores and risk types of all account nodes in the business relationship diagram; based on the risk scores and risk types of all account nodes in the business relationship diagram, selects target account nodes whose risk scores exceed the preset risk control value and whose risk type is the target control type from the business relationship diagram; and sends the business account information corresponding to all target account nodes to the preset risk control management terminal. The business risk control method is applied to financial or banking transaction scenarios, which can identify transaction risks and avoid economic losses of customers; at the same time, the business risk control method can also screen out risk accounts in unspecified accounts, and timely control and handle risk accounts, assisting financial regulatory authorities in screening financial crime cases, which is more intelligent and automated, combined with artificial intelligence machine learning methods, and improves risk identification efficiency and accuracy.

[0170] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through computer-readable instructions, and the computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the aforementioned storage medium can be a non-volatile storage medium such as a disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0171] It should be understood that, although the steps in the flowchart of the accompanying drawings are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.

[0172] To solve the above technical problems, the present application also provides a computer device. Fig.10 , Fig.10 This is a basic structural block diagram of the computer device in this embodiment.

[0173] The computer device 10 includes a memory 10a, a processor 10b, and a network interface 10c which are interconnected and communicated via a system bus. It should be noted that: Fig.10 Only a computer device 10 having a component memory 10a, a processor 10b, and a network interface 10c is shown, but it should be understood that it is not required to implement all the components shown, and more or fewer components may be implemented instead. Among them, those skilled in the art can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to a microprocessor, an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), a digital processor (DSP), an embedded device, etc.

[0174] The computer device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The computer device may interact with a user through a keyboard, a mouse, a remote controller, a touch pad, or a voice control device.

[0175] The memory 10a 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, magnetic disk, optical disk, etc. In some embodiments, the memory 10a can be an internal storage unit of the computer device 10, such as a hard disk or memory of the computer device 10. In other embodiments, the memory 10a can also be an external storage device of the computer device 10, such as a plug-in hard disk equipped on the computer device 10, a smart memory card (Smart Med ia Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. Of course, the memory 10a can also include both the internal storage unit of the computer device 10 and its external storage device. In this embodiment, the memory 10a is generally used to store an operating system and various application software installed on the computer device 10, such as a computer-readable instruction of a business risk management method. In addition, the memory 10a can also be used to temporarily store various data that have been output or are to be output.

[0176] The processor 10b may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chip in some embodiments. The processor 10b is generally used to control the overall operation of the computer device 10. In this embodiment, the processor 10b is used to run the computer-readable instructions stored in the memory 10a or process data, such as computer-readable instructions for running the business risk management method.

[0177] The network interface 10c may include a wireless network interface or a wired network interface. The network interface 10c is generally used to establish a communication connection between the computer device 10 and other electronic devices.

[0178] The computer device proposed in this embodiment belongs to the field of business security supervision technology and is applied to financial business risk management scenarios. This application obtains the data to be detected; extracts features from the data to be detected, extracts business transaction features and business attribute features; inputs the business transaction features and business attribute features into the pre-trained business risk characterization model, obtains the business relationship diagram output by the business risk characterization model and the risk scores and risk types of all account nodes in the business relationship diagram; according to the risk scores and risk types of all account nodes in the business relationship diagram, selects the target account nodes whose risk scores exceed the preset risk control value and whose risk type is the target control type from the business relationship diagram; sends the business account information corresponding to all target account nodes to the preset risk control management terminal. The business risk management method is applied to financial or bank transaction scenarios, which can identify transaction risks and avoid economic losses of customers; at the same time, the business risk management method can also screen out risk accounts in unspecified accounts, and timely manage and handle risk accounts, assist financial regulatory departments in screening financial crime cases, and is more intelligent and automated, combined with artificial intelligence machine learning methods, to improve risk identification efficiency and accuracy.

[0179] The present application also provides another implementation, namely, providing a computer-readable storage medium, wherein the computer-readable storage medium stores computer-readable instructions, and the computer-readable instructions can be executed by a processor to enable the processor to perform the steps of the business risk management method as described above.

[0180] The computer-readable storage medium proposed in this embodiment belongs to the field of business security supervision technology and is applied to financial business risk management scenarios. This application obtains the data to be detected; extracts features from the data to be detected, extracts business transaction features and business attribute features; inputs the business transaction features and business attribute features into the pre-trained business risk characterization model, obtains the business relationship diagram output by the business risk characterization model and the risk scores and risk types of all account nodes in the business relationship diagram; according to the risk scores and risk types of all account nodes in the business relationship diagram, selects the target account nodes whose risk scores exceed the preset risk control value and whose risk type is the target control type from the business relationship diagram; sends the business account information corresponding to all target account nodes to the preset risk control management terminal. The business risk management method is applied to financial or bank transaction scenarios, which can identify transaction risks and avoid economic losses of customers; at the same time, the business risk management method can also screen out risk accounts in unspecified accounts, and timely manage and handle risk accounts, assist financial regulatory departments in screening financial crime cases, and is more intelligent and automated, combined with artificial intelligence machine learning methods, to improve risk identification efficiency and accuracy.

[0181] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0182] Obviously, the embodiments described above are only some embodiments of the present application, rather than all embodiments. The preferred embodiments of the present application are given in the accompanying drawings, but they do not limit the patent scope of the present application. The present application can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present application more thorough and comprehensive. Although the present application is described in detail with reference to the aforementioned embodiments, for those skilled in the art, it is still possible to modify the technical solutions recorded in the aforementioned specific implementation methods, or to perform equivalent replacement of some of the technical features therein. Any equivalent structure made using the contents of the specification and drawings of this application, directly or indirectly used in other related technical fields, is similarly within the scope of patent protection of this application.

Claims

1. A business risk management method, characterized in that: The steps include: Acquire the data to be detected, wherein the data to be detected includes business relationship data of the batch accounts to be detected; Performing feature extraction on the data to be detected, extracting business transaction features and business attribute features, wherein the business transaction features include business flow information between the batch accounts to be detected, and the business attribute features include business account information, business amount information, business frequency information, and business time information of the batch accounts to be detected; Input the business transaction characteristics and the business attribute characteristics into a pre-trained business risk characterization model to obtain a business relationship graph output by the business risk characterization model, as well as risk scores and risk types of all account nodes in the business relationship graph; According to the risk scores and risk types of all account nodes in the business relationship graph, target account nodes whose risk scores exceed a preset risk control value and whose risk types are target control types are screened out from the business relationship graph; Send the business account information corresponding to all target account nodes to the preset risk control management terminal.

2. The business risk management method according to claim 1, characterized in that: The business risk characterization model includes a risk feature identification component, a business relationship graph construction component and a risk type identification component. Before executing the step of inputting the business transaction features and the business attribute features into the pre-trained business risk characterization model to obtain the business relationship graph output by the business risk characterization model, and the risk scores and risk types of all account nodes in the business relationship graph, the method further includes: Acquire a training data set, wherein the training data set includes positive and negative sample data, the positive sample data includes business relationship data of a batch of labeled risky accounts, and the negative sample data includes business relationship data of a batch of labeled non-risky accounts; Performing feature extraction on the training data set to obtain business attribute features contained in all positive and negative sample data and business transaction features contained in all positive and negative sample data; Inputting the business attribute features contained in all the positive and negative sample data into the risk feature identification component to be trained for learning and training, and learning to obtain a trained risk feature identification component; Inputting the business attribute features contained in all the positive and negative sample data and the business transaction features contained in all the positive and negative sample data into the business relationship graph construction component, and obtaining the business transaction relationship graph constructed by the business relationship graph construction component according to the training data set; The business attribute features contained in all the positive and negative sample data and the business transaction relationship diagram are input into the risk type identification component to be trained, and the risk type identification component to be trained is trained to obtain a trained risk type identification component.

3. The business risk management method according to claim 2, characterized in that: The step of inputting the business attribute features contained in all the positive and negative sample data into the risk feature identification component to be trained for learning and training, and learning to obtain the trained risk feature identification component specifically includes: According to the positive and negative annotation knowledge of samples, the business attribute features contained in all positive and negative sample data are sorted out, and the business attribute features corresponding to the positive sample data and the business attribute features corresponding to the negative sample data are sorted out respectively; By comparing the business attribute features corresponding to the positive sample data with the business attribute features corresponding to the negative sample data, the detection features for risk identification are learned; According to the feature data mapping of all detection features in the positive sample data, all detection features are sorted by importance, and the most important detection features with a preset ratio are selected; Setting risk feature detection nodes for the risk feature identification component based on the most important detection features of the preset ratio, and setting risk detection weights for all risk feature detection nodes according to the importance ranking and a preset weight allocation strategy; The risk feature identification component in which the risk feature detection nodes and the risk detection weights have been set is used as the trained risk feature identification component.

4. The business risk management method according to claim 3, characterized in that: The step of ranking all detection features by importance according to the feature data mapping of all detection features in the positive sample data, and screening out the most important detection features with a preset proportion, specifically includes: Count the number of samples corresponding to all detection features in the positive sample data; According to the relationship of sample quantity from high to low, all the detection features are sorted in descending order to obtain the importance sorting result; According to a preset ratio value, extracting detection features of a target ratio from the importance sorting result as the most important detection features of the preset ratio; The step of setting a risk feature detection node for the risk feature identification component based on the most important detection feature of the preset ratio, and setting a risk detection weight for all risk feature detection nodes according to the importance ranking and a preset weight allocation strategy, specifically includes: Setting the most important detection features of the preset proportion as risk feature detection nodes of the risk feature identification component; According to the sequence information of the most important detection features of the preset proportion in the importance ranking result, risk detection weights are set for all risk feature detection nodes, wherein the preset weight allocation strategy includes that the higher the sequence information of the detection feature in the importance ranking result, the greater the risk detection weight corresponding to the detection feature.

5. The business risk management method according to claim 2, characterized in that: The step of inputting the business attribute features contained in all the positive and negative sample data and the business transaction features contained in all the positive and negative sample data into the business relationship graph construction component to obtain the business transaction relationship graph constructed by the business relationship graph construction component according to the training data set specifically includes: Obtaining account information corresponding to all positive and negative sample data respectively from the business attribute features contained in all the positive and negative sample data; The account information corresponding to all the positive and negative sample data is used as a graph node of the business transaction relationship graph; Constructing edge relationships between graph nodes in the business transaction relationship graph according to the business transaction features contained in all the positive and negative sample data; The business transaction relationship graph is generated based on the graph nodes of the business transaction relationship graph and the edge relationships between the graph nodes in the business transaction relationship graph.

6. The business risk management method according to claim 2, characterized in that: The step of inputting the business attribute features and the business transaction relationship diagram contained in all the positive and negative sample data into the risk type identification component to be trained, performing learning training on the risk type identification component to be trained, and obtaining the trained risk type identification component specifically includes: Based on the risk type annotation knowledge, identifying the non-group risk business accounts and the group risk business accounts in the business transaction relationship diagram; Summarize and organize the business attribute characteristics of non-group risk business accounts to obtain summary results of the business attribute characteristics of non-group risk business accounts; Summarize and organize the business attribute characteristics of the group-type risk business accounts to obtain the summary results of the business attribute characteristics of the group-type risk business accounts; According to the summary result of the business attribute characteristics of the non-group risk business account, the type identification characteristics of the non-group risk business account are determined as the first type identification characteristics; According to the summary results of the business attribute characteristics of the group-type risk business account, the type identification characteristics of the group-type risk business account are determined as the second type identification characteristics; According to different risk types, respectively set type recognition weights for the first type recognition feature and the second type recognition feature, and set type feature recognition nodes for the risk type recognition component to be trained according to the first type recognition feature and the second type recognition feature; The risk type identification component in which the type feature identification node and the type identification weight are set is used as the trained risk type identification component.

7. The business risk management method according to claim 2, characterized in that: The step of inputting the business transaction characteristics and the business attribute characteristics into the pre-trained business risk characterization model to obtain the business relationship graph output by the business risk characterization model, and the risk scores and risk types of all account nodes in the business relationship graph, specifically includes: The risk feature identification component performs risk feature detection on the business attribute features of all accounts to be detected, and identifies the risk feature detection nodes hit by all accounts to be detected; The risk feature detection nodes hit by the same account to be detected are aggregated and sorted to obtain the detection node sets corresponding to all accounts to be detected; According to the detection node sets corresponding to all the accounts to be detected and the risk detection weights corresponding to the risk feature detection nodes, the risk detection weights corresponding to the elements in the same detection node set are weighted and summed to obtain the risk detection scores corresponding to all the accounts to be detected; The risk detection scores corresponding to all the accounts to be detected are used as the risk scores corresponding to all the accounts to be detected; Obtaining business account information of all accounts to be detected from the business attribute characteristics; Using the business account information of all the accounts to be detected as graph nodes of the business relationship graph; Identify the edge relationships between all graph nodes based on the business transaction characteristics; Generate the business relationship graph based on the graph node and the edge relationships between all graph nodes; The risk type identification component identifies the risk type of the business attribute characteristics of all accounts to be detected, and identifies the type feature identification nodes hit by all accounts to be detected; The type feature identification nodes hit by the same account to be detected are collected and sorted to obtain the type node sets corresponding to all the accounts to be detected; According to the type node sets corresponding to all the accounts to be detected and the type identification weights corresponding to the type feature identification nodes, the type identification weights corresponding to the elements in the same type node set are weighted summed to obtain the type identification scores corresponding to all the accounts to be detected; Determine the risk types corresponding to all accounts to be detected according to the type identification score and the preset type identification rules; According to the business account information of all the accounts to be detected, the risk scores and risk types corresponding to all the accounts to be detected are marked at the corresponding graph node positions in the business relationship graph to obtain a marked business relationship graph; The annotated business relationship diagram is outputted through the business risk characterization model.

8. A business risk management and control device, characterized in that: include: A module for acquiring data to be detected, used to acquire data to be detected, wherein the data to be detected includes business relationship data of batch accounts to be detected; A feature extraction module is used to extract features from the data to be detected, and extract business transaction features and business attribute features, wherein the business transaction features include business flow information between the batch accounts to be detected, and the business attribute features include business account information, business amount information, business frequency information, and business time information of the batch accounts to be detected; A model processing module, used to input the business transaction characteristics and the business attribute characteristics into a pre-trained business risk characterization model, obtain a business relationship graph output by the business risk characterization model, and risk scores and risk types of all account nodes in the business relationship graph; A target account node screening module, configured to screen target account nodes whose risk scores exceed a preset risk control value and whose risk types are target control types from the business relationship graph according to the risk scores and risk types of all account nodes in the business relationship graph; The business account information sending module is used to send the business account information corresponding to all target account nodes to the preset risk control management terminal.

9. A computer device, characterized in that: It includes a memory and a processor, the memory stores computer-readable instructions, and the processor implements the steps of the business risk management method as described in any one of claims 1 to 7 when executing the computer-readable instructions.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the steps of the business risk management method according to any one of claims 1 to 7 are implemented.

Images