Suspicious right activity identification method and device

By applying machine learning technology in the bank equity activity audit system, extracting feature information and initiator information, training suspicious equity activity recognition model, identifying suspicious behavior, the problems of low identification efficiency and poor accuracy in the existing technology are solved, and the problem of low recognition efficiency and poor accuracy is achieved Efficient, accurate identification and risk reduction of bank equity activities.

CN119090614APending Publication Date: 2024-12-06CHINA CONSTRUCTION BANK +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411221482.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-02
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

The prior art is inefficient and poorly accurate in identifying suspicious behaviors in bank equity activities, and cannot effectively prevent corruption.

Method used

Using machine learning technology, by extracting characteristic information and initiator information from the equity activity audit system, training suspicious equity activity identification models, identifying equity activities that may involve corruption, and improving the accuracy and efficiency of the audit.

Benefits of technology

Accurate and efficient identification of suspicious rights and interests activities has been achieved, the accuracy and efficiency of audits have been improved, and the risk of corruption has been reduced.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119090614A_ABST
    Figure CN119090614A_ABST
Patent Text Reader

Abstract

The invention discloses a suspicious right and interest activity identification method and device, relates to the field of artificial intelligence and can also be used for the financial field, and the method comprises the steps: obtaining right and interest activity auditing information of a target right and interest activity; obtaining right and interest activity information and activity initiator information of the target right and interest activity; extracting right and interest activity auditing feature information from the right and interest activity auditing information, and extracting activity and initiator feature information from the right and interest activity information and the activity initiator information; and determining an identification result of the target right and interest activity according to the right and interest activity auditing feature information, the activity and initiator feature information and a trained suspicious right and interest activity identification model. According to the invention, the identification efficiency and accuracy of suspicious right activities can be improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and more specifically, to a method and device for identifying suspicious equity activities. Background Art

[0002] Equity activities are a type of marketing activities of banks, including issuing instant discounts, vouchers, video memberships, physical objects and other benefits to customers. The initiators of equity activities can be various departments of the bank, lower-level branches, etc., which usually need to be reviewed by the department / branch and finally reviewed and finally reviewed by the competent department of the equity activities, involving multiple links of review by different departments and different positions. Equity activities involve valuable items and may involve corruption. Some people apply for equity activities to profit from them. The existing technology requires a dedicated person to review and identify corruption issues such as equity activities for the purpose of profit, and discover suspicious equity activities. However, the current manual review and identification scheme has the problems of low efficiency and poor accuracy. The existing technology urgently needs a more efficient and accurate scheme to identify suspicious equity activities. Summary of the invention

[0003] In order to solve at least one technical problem in the above-mentioned background technology, the present invention proposes a method and device for identifying suspicious equity activities.

[0004] In order to achieve the above object, according to one aspect of the present invention, a method for identifying suspicious equity activities is provided, the method comprising:

[0005] Acquire the equity activity review information of the target equity activity, wherein the target equity activity is initiated for review by the activity initiator in the equity activity review system and is reviewed through the equity activity review process set in the equity activity review system, and the equity activity review information is obtained from the equity activity review system;

[0006] Obtain the equity activity information of the target equity activity and the information of the activity initiator;

[0007] Extracting rights and interests activity audit feature information from the rights and interests activity audit information, and extracting activity and initiator feature information from the rights and interests activity information and the activity initiator information;

[0008] The identification result of the target equity activity is determined based on the equity activity review characteristic information, the activity and initiator characteristic information and the trained suspicious equity activity identification model, wherein the suspicious equity activity identification model is generated by generating training samples based on the equity activity review information, equity activity information and activity initiator information of multiple equity activities that have been reviewed and approved by the equity activity review system in the past, and then training a preset machine learning model based on the training samples.

[0009] Optionally, the method for identifying suspicious equity activities further includes:

[0010] Obtaining equity activity review information, equity activity information, and activity initiator information of multiple equity activities that have been reviewed and approved by the equity activity review system in the past, and extracting equity activity review feature information for model training and activity and initiator feature information for model training;

[0011] Generate model input data for model training according to the equity activity review feature information used for model training and the activity and initiator feature information used for model training, and then set the recognition result label corresponding to the model input data used for model training to obtain a training sample;

[0012] The preset machine learning model is trained according to the training samples to obtain the suspicious equity activity identification model.

[0013] Optionally, determining the identification result of the target equity activity according to the equity activity review feature information, the activity and initiator feature information, and a trained suspicious equity activity identification model includes:

[0014] Generate model input data based on the equity activity review characteristic information and the activity and initiator characteristic information;

[0015] The model input data is input into the suspicious equity activity identification model to obtain the identification result of the target equity activity output by the suspicious equity activity identification model.

[0016] Optionally, the generating model input data according to the equity activity review characteristic information and the activity and initiator characteristic information includes:

[0017] Standardize the review characteristic information of the equity activities and the characteristic information of the activities and initiators;

[0018] Based on the weights obtained through iterative model training, weighted summation is performed on the processed equity activity review feature information and the activity and initiator feature information to obtain fused feature information;

[0019] A variational autoencoder is used to convert the fused feature information into a feature vector of a preset dimension to obtain the model input data.

[0020] Optionally, the identification result includes: a first result for indicating suspicious equity activities and a second result for indicating normal equity activities;

[0021] The method further comprises:

[0022] If the identification result is the first result, the information of the reviewer in the review link of the competent department and the information of the final reviewer in the final review link of the competent department in the rights and interests activity review process corresponding to the target rights and interests activity are obtained from the rights and interests activity review system, and then a suspicious rights and interests activity warning is sent to the reviewer and the final reviewer based on this information to remind the reviewer and the final reviewer to re-review the target rights and interests activity.

[0023] Optionally, the rights and interests activity review system is used to review the rights and interests activities applied for by the activity initiators, and only the rights and interests activities that have been reviewed and approved by the rights and interests activity review system can be put online; the rights and interests activity review system contains corresponding rights and interests activity review processes set up for different types of activity initiators; each rights and interests activity review process includes multiple links connected in sequence, and each link is reviewed in sequence, and each link corresponds to a different review role.

[0024] Optionally, the rights and interests activity review characteristic information includes: the number of links in the activity review process, the total processing time of the activity review process, the average processing time of each link in the activity review process, whether the review link of the competent department in the activity review process has been returned, whether the final review link of the competent department in the activity review process has been returned, whether there are attachments in the link of initiating rights and interests activities in the activity review process, the department type corresponding to the link of initiating rights and interests activities in the activity review process, and whether the activity review process has been withdrawn.

[0025] Optionally, the activity and initiator characteristic information includes: the number of equity activities initiated by the activity initiator within a preset time range and the total amount of equity, the number of equity activities initiated by the operator who initiates the equity activities within a preset time range and the total amount of equity, the initiation time of the equity activity, the start and end time range of the equity activity, the type of equity product, the total inventory of equity products, the restricted number of equity products per person, the clarity of product description, the failure rate of previous awarding of equity products provided by the same supplier, whether the activity involves jumping to a third-party link, whether the equity activity restricts participating customers, and at least one of the number of customers restricted by the equity activity.

[0026] In order to achieve the above object, according to another aspect of the present invention, a suspicious equity activity identification device is provided, the device comprising:

[0027] A rights and interests activity audit information acquisition unit, used to acquire rights and interests activity audit information of a target rights and interests activity, wherein the target rights and interests activity is initiated for audit by an activity initiator in a rights and interests activity audit system and audited through a rights and interests activity audit process set in the rights and interests activity audit system, and the rights and interests activity audit information is obtained from the rights and interests activity audit system;

[0028] A rights and interests activity and initiator information acquisition unit, used to acquire rights and interests activity information and activity initiator information of a target rights and interests activity;

[0029] A feature extraction unit, configured to extract rights and interests activity review feature information from the rights and interests activity review information, and to extract activity and initiator feature information from the rights and interests activity information and the activity initiator information;

[0030] An identification unit is used to determine the identification result of the target equity activity based on the equity activity review feature information, the activity and initiator feature information and a trained suspicious equity activity identification model, wherein the suspicious equity activity identification model is generated by generating training samples based on the equity activity review information, equity activity information and activity initiator information of multiple equity activities that have been reviewed and approved by the equity activity review system in the past, and then training a preset machine learning model based on the training samples.

[0031] In order to achieve the above-mentioned purpose, according to another aspect of the present invention, a computer device is also provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned method for identifying suspicious equity activities when executing the computer program.

[0032] In order to achieve the above-mentioned purpose, according to another aspect of the present invention, a computer-readable storage medium is provided, on which a computer program / instruction is stored, and when the computer program / instruction is executed by a processor, the steps of the above-mentioned suspicious equity activity identification method are implemented.

[0033] In order to achieve the above-mentioned purpose, according to another aspect of the present invention, a computer program product is also provided, including a computer program / instruction, which implements the steps of the above-mentioned suspicious equity activity identification method when executed by a processor.

[0034] The beneficial effects of the present invention are:

[0035] The embodiment of the present invention extracts equity activity audit feature information from the equity activity audit information of the target equity activity, extracts activity and initiator feature information from the equity activity information and activity initiator information of the target equity activity, and then determines the identification result of the target equity activity based on the equity activity audit feature information, the activity and initiator feature information and the trained suspicious equity activity identification model, thereby achieving the technical effect of accurately and efficiently identifying suspicious equity activities. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0037] Figure 1 is a flow chart of a method for identifying suspicious equity activities according to an embodiment of the present invention;

[0038] Figure 2 is a flow chart of model training according to an embodiment of the present invention;

[0039] Figure 3 is a flow chart of model identification processing according to an embodiment of the present invention;

[0040] Figure 4 is a flow chart of generating model input data according to an embodiment of the present invention;

[0041] Figure 5 This is a schematic diagram of the rights and interests activity review process of an embodiment of the present invention;

[0042] Figure 6 This is a schematic diagram of an example of a rights and interests activity review process according to an embodiment of the present invention;

[0043] Figure 7 This is a first structural block diagram of a device for identifying suspicious equity activities according to an embodiment of the present invention;

[0044] Figure 8 This is a second structural block diagram of the device for identifying suspicious equity activities according to an embodiment of the present invention;

[0045] Fig. 9 It is a schematic diagram of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0046] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings 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 should fall within the scope of protection of the present invention.

[0047] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0048] It should be noted that the terms "including" and "having" and any variations thereof in the specification and claims of the present invention and the above-mentioned drawings are intended to cover non-exclusive inclusions. For example, a process, method, system, product or apparatus comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or apparatus.

[0049] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0050] The information collected in the technical solution of this application is information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of relevant data comply with the relevant laws, regulations and standards of relevant countries and regions, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0051] The acquisition, transmission, storage, use, and processing of data in the technical solution of this application comply with the relevant provisions of national laws and regulations.

[0052] It should be noted that in the embodiments of the present application, certain software, components, models and other existing solutions in the industry may be mentioned, and they should be regarded as exemplary. Their purpose is only to illustrate the feasibility of implementing the technical solution of the present application, but it does not mean that the applicant has or will necessarily use the solution.

[0053] It should be noted that the suspicious equity activity identification method and device of the present invention can be used in the financial field, and can also be used in any field other than the financial field. The application field of the suspicious equity activity identification method and device of the present invention is not limited.

[0054] It should be noted that the suspicious equity activity identification method of the present invention is aimed at equity activities that have been approved by the equity activity review system. The equity activity review system of the present invention is a business flow system for approving equity activities. Usually, when the activity initiator wants to apply for an equity activity, it is necessary to initiate the approval of the equity activity in the equity activity review system. The equity activity can only go online after it has been approved by the equity activity review system.

[0055] Equity activities involve valuables and may involve corruption. Someone in the enterprise may apply for excessive equity activities or set special conditions for equity activities, thereby obtaining a large number of equity items and then reselling them for profit, causing losses to the enterprise. The current equity activity audit system is unable to audit and identify such corruption problems for the purpose of profit. Even after the multi-link approval of the equity activity audit system, the approved equity activities cannot be guaranteed not to involve corruption. At present, it is necessary to set up auditors to further audit and identify the equity activities approved by the equity activity audit system based on expert experience, but the current manual audit and identification scheme has the problems of low efficiency and poor accuracy. In order to solve this problem, the present invention proposes a suspicious equity activity identification method based on machine learning, which can accurately and efficiently identify suspicious equity activities. Suspicious equity activities refer to equity activities that may involve corruption problems. Then, auditors can audit suspicious equity activities more specifically based on the identification results, thereby improving the efficiency and accuracy of audit identification.

[0056] Figure 1 is a flow chart of a method for identifying suspicious equity activities according to an embodiment of the present invention. Figure 1 As shown, in one embodiment of the present invention, the method for identifying suspicious equity activities of the present invention includes steps S101 to S104.

[0057] Step S101, obtaining the equity activity audit information of the target equity activity, wherein the target equity activity is initiated for audit by the activity initiator in the equity activity audit system, and the audit is completed through the equity activity audit process set in the equity activity audit system, and the equity activity audit information is obtained from the equity activity audit system.

[0058] In one embodiment of the present invention, the rights and interests activity review system is used to review the rights and interests activities applied for by the activity initiator, and only the rights and interests activities that have been reviewed and approved by the rights and interests activity review system can be put online; the rights and interests activity review system contains corresponding rights and interests activity review processes set for different types of activity initiators; each rights and interests activity review process includes multiple links connected in sequence, each link is reviewed in sequence, and each link corresponds to a different review role.

[0059] Figure 5Schematic diagram of the rights and interests activity review process of the embodiment of the present invention. Figure 5 As shown, the rights and interests activity review process of the present invention can go through multiple links, and processes initiated by different positions and different users can start at different links. Except for the end link, each link points to a next link. The link can be freely configured, and the next link can also be freely modified. Even if the link is modified during the flow of the process instance, it can take effect immediately.

[0060] When the rights and interests activity review process is actually initiated, create a rights and interests activity review process instance, such as Figure 6 As shown in the figure. In a process instance, a link instance is created every time a link is passed.

[0061] The suspicious equity activity identification method of the present invention is aimed at the equity activities that have been approved by the equity activity review system. Therefore, it is necessary to first obtain the equity activity review information of the target equity activity from the equity activity review system. The equity activity review information specifically records the detailed review of each link in the equity activity review process, for example, the processing results of each link, the processing time of each link, the attachments of each link, the operator information of each link, etc.

[0062] Step S102, obtaining the equity activity information of the target equity activity and the activity initiator information.

[0063] In order to improve the accuracy of identifying suspicious equity activities, the present invention also considers the information of the equity activities themselves and the information of the initiators of the equity activities to participate in model training and identification, which helps to improve the accuracy of model identification.

[0064] Equity activity information may include: the initiation time of the equity activity, the start and end time range of the equity activity, the type of equity products, the total inventory of equity products, the limited number of equity products per person, the clarity of product descriptions, the failure rate of previous awards for equity products provided by the same supplier, whether the activity involves jumping to a third-party link, whether the equity activity restricts participating customers, the number of participating customers restricted by the equity activity, etc.

[0065] The activity initiator information may include: the department type of the activity initiator, the number and total amount of equity activities initiated by the activity initiator within the preset time range, the number and total amount of equity activities initiated by the operator who initiated the equity activities within the preset time range, the collection rate of equity activities initiated by the activity initiator in history, etc.

[0066] Step S103: extracting rights and interests activity review characteristic information from the rights and interests activity review information, and extracting activity and initiator characteristic information from the rights and interests activity information and the activity initiator information.

[0067] In one embodiment of the present invention, the present invention adopts a feature screening method to screen out features that are significant for identification from equity activity review information, equity activity information and activity initiator information. Feature screening can adopt a variety of existing methods, for example, principal component analysis, screening based on linear regression, random forest model, etc.

[0068] In one embodiment of the present invention, the rights and interests activity review characteristic information includes: the number of links in the activity review process, the total processing time of the activity review process, the average processing time of each link in the activity review process, whether the review link of the competent department in the activity review process has been returned, whether the final review link of the competent department in the activity review process has been returned, whether there are attachments in the link of initiating rights and interests activities in the activity review process, the department type corresponding to the link of initiating rights and interests activities in the activity review process, and whether the activity review process has been withdrawn.

[0069] In one embodiment of the present invention, the activity and initiator characteristic information includes: the number of equity activities initiated by the activity initiator within a preset time range and the total amount of equity, the number of equity activities initiated by the operator who initiates the equity activity within a preset time range and the total amount of equity, the initiation time of the equity activity, the start and end time range of the equity activity, the type of equity product, the total inventory of equity products, the limited number of equity products per person, the clarity of the product description, the failure rate of previous awarding of equity products provided by the same supplier, whether the activity involves jumping to a third-party link, whether the equity activity restricts participating customers, and at least one of the number of participating customers restricted by the equity activity.

[0070] Step S104, determining the identification result of the target equity activity based on the equity activity review characteristic information, the activity and initiator characteristic information and the trained suspicious equity activity identification model, wherein the suspicious equity activity identification model is generated by generating training samples based on the equity activity review information, equity activity information and activity initiator information of multiple equity activities that have been reviewed and approved by the equity activity review system in the past, and then training a preset machine learning model based on the training samples.

[0071] The present invention uses artificial intelligence technology to identify suspicious equity activities through a trained suspicious equity activity identification model, achieving the beneficial effect of accurately and efficiently identifying suspicious equity activities, and helping to improve the accuracy and efficiency of identifying suspicious equity activities. The present invention combines the audit information of equity activities in the equity activity audit system, the information of the equity activities themselves, and the information of the initiator of the equity activities, and comprehensively identifies them, making the identification results more accurate, thereby improving the accuracy and credibility of identifying suspicious equity activities.

[0072] Combine the following Figure 5and Figure 6 A brief description of the review of equity activities through the equity activity review system.

[0073] 1. Create / modify equity activities

[0074] According to the creator's position and process type, query the process start link in the equity activity process link information table;

[0075] Add a new equity activity process instance with the status of Initial;

[0076] Add a new link instance information (start link 1, status is initial);

[0077] Insert a piece of information into the side table.

[0078] 2. The creator submits the review

[0079] The status of the equity activity process instance is changed to under review;

[0080] Current stage instance: Start stage 1 changes to processed status;

[0081] Query the next link in the rights and interests activity process and create a new link instance information (review link 2, status is pending).

[0082] 3.1. Approved by the auditor

[0083] Current stage example: Review stage 2 is changed to processed;

[0084] Query the next link in the rights and interests activity process and create a new link instance information (review link 20, status is pending).

[0085] 3.2.Rejection by the Reviewer

[0086] The status of the equity activity process instance is changed to return;

[0087] Current stage example: Review stage 2 is changed to returned;

[0088] Go back to the start node and add a new link instance information (start link 1, the status is returned and pending).

[0089] 4.1. Reviewer approved

[0090] Query the next stage of the equity activity process, the current stage example: the review stage 20 is changed to processed;

[0091] Query the next link in the rights and interests activity process and create a new link instance information (final review link 30, status is pending).

[0092] 4.2.Return by the Reviewer

[0093] The status of the equity activity process instance is changed to return;

[0094] Current stage example: Review stage 20 is changed to returned;

[0095] Go back to the start node and add a new link instance information (start link 1, the status is returned and pending).

[0096] 5.1. Approval by the final reviewer

[0097] The status of the equity activity process instance is changed to approved;

[0098] Current stage example: Review stage 30 is changed to processed;

[0099] Query the equity activity process information table and find that it has reached the end node. The review process is completed and the secondary table information is inserted / updated into the original table.

[0100] 5.2. Return by the final judge

[0101] The status of the equity activity process instance is changed to return;

[0102] Current stage example: Review stage 30 is changed to returned;

[0103] Go back to the start node and create a new link instance information (start link 1, the status is returned and pending).

[0104] 6. Withdrawal

[0105] When the current link is pending, if the previous link is the initiator, withdrawal is allowed;

[0106] The equity activity process instance is changed to initial;

[0107] Current stage example: The current stage is changed to withdraw;

[0108] Create a new link instance information (start link 1, status is initial).

[0109] 7. Abandonment

[0110] Abandonment is allowed only when the status of the equity activity process instance is initial or returned;

[0111] The equity activity process instance is changed to abandoned;

[0112] Current stage instance: The current stage is changed to obsolete.

[0113] Figure 2 is a flow chart of the model training of an embodiment of the present invention, such as Figure 2As shown, in one embodiment of the present invention, the process of training a suspicious equity activity identification model of the present invention includes steps S201 to S203.

[0114] Step S201, obtain the equity activity review information, equity activity information and activity initiator information of multiple equity activities that have been reviewed and approved by the equity activity review system in the past, and extract the equity activity review feature information used for model training and the activity and initiator feature information used for model training.

[0115] In one embodiment of the present invention, the auditor audits and identifies the rights and interests activities that have been approved by the rights and interests activity audit system in the past based on expert experience, and confirms the rights and interests activities that do not involve corruption issues as negative samples and the rights and interests activities that are determined to involve corruption issues as positive samples, thereby creating training samples.

[0116] Step S202, generating model input data for model training based on the equity activity review feature information used for model training and the activity and initiator feature information used for model training, and then setting the recognition result label corresponding to the model input data for model training to obtain a training sample.

[0117] In one embodiment of the present invention, the training sample includes: model input data for model training and recognition result labels.

[0118] Step S203: training a preset machine learning model according to the training samples to obtain the suspicious equity activity identification model.

[0119] In one embodiment of the present invention, the machine learning model may adopt classification models such as decision trees, random forests, support vector machines, etc., or may adopt a neural network model.

[0120] The recognition result of the suspicious equity activity recognition model of the present invention represents the probability of a suspicious equity activity under the current input data, and the specific recognition result is a probability value. The present invention compares the recognition result of the suspicious equity activity recognition model with a preset probability threshold. If the recognition result is greater than the preset probability threshold, it is determined to be a suspicious equity activity, otherwise it is determined to be a normal equity activity.

[0121] It can be seen that the present invention uses training data and artificial intelligence algorithms to perform model training, constructs a model that can accurately identify suspicious equity activities, and improves the accuracy of performance status prediction.

[0122] Figure 3 is a flow chart of the model recognition process according to an embodiment of the present invention, such as Figure 3As shown, in one embodiment of the present invention, the above step S104 determines the identification result of the target equity activity based on the equity activity review feature information, the activity and initiator feature information and the trained suspicious equity activity identification model, specifically including step S301 and step S302.

[0123] Step S301, generating model input data according to the equity activity review characteristic information and the activity and initiator characteristic information.

[0124] In one embodiment of the present invention, this step can merge the rights and interests activity review characteristic information and the activity and initiator characteristic information to obtain model input data. When merging, the rights and interests activity review characteristic information and the activity and initiator characteristic information can be converted into data of the same dimension, and then weighted calculation is performed based on the weights set by experts to obtain model input data.

[0125] In another embodiment of the present invention, this step can use the equity activity review characteristic information and the activity and initiator characteristic information as model input data and input them into the model at the same time. At this time, the model can adopt a neural network model containing multiple input layers.

[0126] Step S302: input the model input data into the suspicious equity activity identification model to obtain the identification result of the target equity activity output by the suspicious equity activity identification model.

[0127] Figure 4 is a flow chart of generating model input data according to an embodiment of the present invention, such as Figure 4 As shown, in one embodiment of the present invention, the above step S301 generates model input data according to the equity activity review characteristic information and the activity and initiator characteristic information, specifically including steps S401 to S403.

[0128] Step S401, standardizing the equity activity review characteristic information and the activity and initiator characteristic information.

[0129] Step S402, based on the weights obtained through iterative model training, weighted sum is performed on the processed equity activity review feature information and the activity and initiator feature information to obtain fused feature information.

[0130] In another embodiment of the present invention, this step obtains weights of the equity activity review feature information and the activity and initiator feature information in advance through iterative model training, and then performs weighted summation based on the iteratively obtained weights.

[0131] In another embodiment of the present invention, this step may also perform weighted summation of the processed rights and interests activity review feature information and the activity and initiator feature information based on the weights set by the experts to obtain the fused feature information.

[0132] Step S403: Using a variational autoencoder to convert the fused feature information into a feature vector of a preset dimension to obtain the model input data.

[0133] In one embodiment of the present invention, the present invention may use a variational autoencoder (VAE) to convert the fused feature information into a feature vector of a preset dimension. The present invention can simplify the model design, improve the training efficiency, and in some cases improve the prediction performance of the model by processing the model input data into a feature vector of a preset dimension.

[0134] In one embodiment of the present invention, the identification result includes: a first result for indicating suspicious equity activities and a second result for indicating normal equity activities.

[0135] In one embodiment of the present invention, the method for identifying suspicious equity activities of the present invention further includes:

[0136] If the identification result is the first result, the information of the reviewer in the review link of the competent department and the information of the final reviewer in the final review link of the competent department in the rights and interests activity review process corresponding to the target rights and interests activity are obtained from the rights and interests activity review system, and then a suspicious rights and interests activity warning is sent to the reviewer and the final reviewer based on this information to remind the reviewer and the final reviewer to re-review the target rights and interests activity.

[0137] After obtaining the identification result, if the result is a suspicious rights and interests activity, the present invention also sends an early warning message to the final reviewer of the final review stage of the competent department and the reviewer of the review stage of the competent department to remind the reviewer and the final reviewer to re-examine the target rights and interests activity.

[0138] It can be seen from the above embodiments that the present invention provides a solution for identifying suspicious equity activities by using artificial intelligence technology through a trained suspicious equity activity identification model, achieving the beneficial effect of accurately and efficiently identifying suspicious equity activities, and helping to improve the accuracy and efficiency of identifying suspicious equity activities. The present invention combines the audit information of equity activities in the equity activity audit system, the information of the equity activities themselves, and the information of the initiator of the equity activities, and comprehensively identifies them, making the identification results more accurate, thereby improving the accuracy and credibility of identifying suspicious equity activities.

[0139] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0140] Based on the same inventive concept, an embodiment of the present invention also provides a suspicious equity activity identification device, which can be used to implement the suspicious equity activity identification method described in the above embodiment, as described in the following embodiment. Since the principle of solving the problem by the suspicious equity activity identification device is similar to that of the suspicious equity activity identification method, the embodiment of the suspicious equity activity identification device can refer to the embodiment of the suspicious equity activity identification method, and the repeated parts will not be repeated. As used below, the term "unit" or "module" can be a combination of software and / or hardware that implements the predetermined function. Although the device described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceived.

[0141] Figure 7 is a first structural block diagram of a device for identifying suspicious equity activities according to an embodiment of the present invention. Figure 7 As shown, in one embodiment of the present invention, the suspicious equity activity identification device of the present invention includes:

[0142] The equity activity audit information acquisition unit 1 is used to acquire the equity activity audit information of the target equity activity, wherein the target equity activity is initiated for audit by the activity initiator in the equity activity audit system and is audited through the equity activity audit process set in the equity activity audit system, and the equity activity audit information is obtained from the equity activity audit system;

[0143] The rights and interests activity and initiator information acquisition unit 2 is used to acquire the rights and interests activity information and the activity initiator information of the target rights and interests activity;

[0144] A feature extraction unit 3 is used to extract the rights and interests activity review feature information from the rights and interests activity review information, and to extract the activity and initiator feature information from the rights and interests activity information and the activity initiator information;

[0145] The identification unit 4 is used to determine the identification result of the target equity activity based on the equity activity review feature information, the activity and initiator feature information and the trained suspicious equity activity identification model, wherein the suspicious equity activity identification model is generated by generating training samples based on the equity activity review information, equity activity information and activity initiator information of multiple equity activities that have been reviewed and approved by the equity activity review system in the past, and then training a preset machine learning model based on the training samples.

[0146] Figure 8 is a second structural block diagram of the device for identifying suspicious equity activities according to an embodiment of the present invention. Figure 8 As shown, in one embodiment of the present invention, the suspicious equity activity identification device of the present invention further includes:

[0147] A historical data acquisition unit 5 is used to acquire the equity activity review information, equity activity information and activity initiator information of multiple equity activities that have been reviewed and approved by the equity activity review system in the past, and extract the equity activity review feature information used for model training and the activity and initiator feature information used for model training;

[0148] A training sample generating unit 6 is used to generate model input data for model training according to the equity activity review characteristic information used for model training and the activity and initiator characteristic information used for model training, and then set the recognition result label corresponding to the model input data used for model training to obtain a training sample;

[0149] The model training unit 7 is used to train the preset machine learning model according to the training samples to obtain the suspicious equity activity identification model.

[0150] In one embodiment of the present invention, the identification unit 4 of the present invention comprises:

[0151] A model input data generation module, used to generate model input data according to the equity activity review characteristic information and the activity and initiator characteristic information;

[0152] The model identification module is used to input the model input data into the suspicious equity activity identification model to obtain the identification result of the target equity activity output by the suspicious equity activity identification model.

[0153] In one embodiment of the present invention, the model input data generating module includes:

[0154] A standardization processing submodule, used for standardizing the equity activity review characteristic information and the activity and initiator characteristic information;

[0155] A feature fusion submodule, for performing weighted summation of the processed equity activity review feature information and the activity and initiator feature information based on the weights obtained through iterative model training to obtain fused feature information;

[0156] The variational autoencoder processing submodule is used to use a variational autoencoder to convert the fused feature information into a feature vector of a preset dimension to obtain the model input data.

[0157] In one embodiment of the present invention, the identification result includes: a first result for indicating suspicious equity activities and a second result for indicating normal equity activities;

[0158] In one embodiment of the present invention, the suspicious equity activity identification device of the present invention further includes:

[0159] An early warning information sending unit is used to obtain, if the identification result is the first result, the information of the reviewer in the review stage of the competent department and the information of the final reviewer in the final review stage of the competent department in the rights and interests activity review process corresponding to the target rights and interests activity from the rights and interests activity review system, and then send suspicious rights and interests activity warnings to the reviewer and the final reviewer based on this information, so as to remind the reviewer and the final reviewer to re-examine the target rights and interests activity.

[0160] In order to achieve the above object, according to another aspect of the present application, a computer device is also provided. Fig. 9 As shown, the computer device includes a memory, a processor, a communication interface and a communication bus. A computer program that can be executed on the processor is stored in the memory. When the processor executes the computer program, the steps in the above-mentioned embodiment method are implemented.

[0161] The processor may be a central processing unit (CPU). The processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or a combination of the above chips.

[0162] The memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs, non-transitory computer executable programs and units, such as the corresponding program units in the above method embodiments of the present invention. The processor executes various functional applications of the processor and works data processing by running the non-transitory software programs, instructions and modules stored in the memory, that is, implementing the method in the above method embodiments.

[0163] The memory may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created by the processor, etc. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory may optionally include a memory remotely arranged relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0164] The one or more units are stored in the memory, and when executed by the processor, the method in the above embodiment is performed.

[0165] The specific details of the above-mentioned computer device can be understood by referring to the corresponding related descriptions and effects in the above-mentioned embodiments, and will not be repeated here.

[0166] In order to achieve the above purpose, according to another aspect of the present application, a computer-readable storage medium is also provided, wherein the computer-readable storage medium stores a computer program, and the computer program implements the steps in the above-mentioned suspicious equity activity identification method when executed in a computer processor. Those skilled in the art can understand that the implementation of all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and the program can include the processes of the embodiments of the above-mentioned methods when executed. Among them, the storage medium can be a disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory (Flash Memory), a hard disk (Hard Disk Drive, abbreviated as: HDD) or a solid-state drive (SSD), etc.; the storage medium can also include a combination of the above-mentioned types of memory.

[0167] In order to achieve the above-mentioned purpose, according to another aspect of the present application, a computer program product is also provided, including a computer program / instruction, which implements the steps of the above-mentioned suspicious equity activity identification method when executed by a processor.

[0168] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, and optionally, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0169] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for identifying suspicious equity activities, It is characterized in that include: Acquire the equity activity review information of the target equity activity, wherein the target equity activity is initiated for review by the activity initiator in the equity activity review system and is reviewed through the equity activity review process set in the equity activity review system, and the equity activity review information is obtained from the equity activity review system; Obtain the equity activity information of the target equity activity and the information of the activity initiator; Extracting rights and interests activity audit feature information from the rights and interests activity audit information, and extracting activity and initiator feature information from the rights and interests activity information and the activity initiator information; The identification result of the target equity activity is determined based on the equity activity review characteristic information, the activity and initiator characteristic information and the trained suspicious equity activity identification model, wherein the suspicious equity activity identification model is generated by generating training samples based on the equity activity review information, equity activity information and activity initiator information of multiple equity activities that have been reviewed and approved by the equity activity review system in the past, and then training a preset machine learning model based on the training samples.

2. The method for identifying suspicious equity activities according to claim 1, It is characterized in that Also includes: Obtaining equity activity review information, equity activity information, and activity initiator information of multiple equity activities that have been reviewed and approved by the equity activity review system in the past, and extracting equity activity review feature information for model training and activity and initiator feature information for model training; Generate model input data for model training according to the equity activity review feature information used for model training and the activity and initiator feature information used for model training, and then set the recognition result label corresponding to the model input data used for model training to obtain a training sample; The preset machine learning model is trained according to the training samples to obtain the suspicious equity activity identification model.

3. The method for identifying suspicious equity activities according to claim 1, It is characterized in that The step of determining the identification result of the target equity activity based on the equity activity review feature information, the activity and initiator feature information, and the trained suspicious equity activity identification model includes: Generate model input data based on the equity activity review characteristic information and the activity and initiator characteristic information; The model input data is input into the suspicious equity activity identification model to obtain the identification result of the target equity activity output by the suspicious equity activity identification model.

4. The method for identifying suspicious equity activities according to claim 3, It is characterized in that The generating model input data according to the equity activity review characteristic information and the activity and initiator characteristic information includes: Standardize the review characteristic information of the equity activities and the characteristic information of the activities and initiators; Based on the weights obtained through iterative model training, weighted summation is performed on the processed equity activity review feature information and the activity and initiator feature information to obtain fused feature information; A variational autoencoder is used to convert the fused feature information into a feature vector of a preset dimension to obtain the model input data.

5. The method for identifying suspicious equity activities according to claim 1, It is characterized in that The identification result includes: a first result for indicating suspicious equity activities and a second result for indicating normal equity activities; The method further comprises: If the identification result is the first result, the information of the reviewer in the review link of the competent department and the information of the final reviewer in the final review link of the competent department in the rights and interests activity review process corresponding to the target rights and interests activity are obtained from the rights and interests activity review system, and then a suspicious rights and interests activity warning is sent to the reviewer and the final reviewer based on this information to remind the reviewer and the final reviewer to re-review the target rights and interests activity.

6. The method for identifying suspicious equity activities according to claim 1, It is characterized in that The rights and interests activity review system is used to review the rights and interests activities applied for by the activity initiator. Only the rights and interests activities that have been reviewed and approved by the rights and interests activity review system can be put online; the rights and interests activity review system contains corresponding rights and interests activity review processes set up for different types of activity initiators; each rights and interests activity review process includes multiple links connected in sequence, and each link is reviewed in sequence, and each link corresponds to a different review role.

7. The method for identifying suspicious equity activities according to claim 1, It is characterized in that The rights and interests activity review characteristic information includes: the number of links in the activity review process, the total processing time of the activity review process, the average processing time of each link in the activity review process, whether the review link of the competent department in the activity review process has been returned, whether the final review link of the competent department in the activity review process has been returned, whether there are attachments in the link of initiating rights and interests activities in the activity review process, the department type corresponding to the link of initiating rights and interests activities in the activity review process, and whether the activity review process has been withdrawn.

8. The method for identifying suspicious equity activities according to claim 1, It is characterized in that The activity and initiator characteristic information includes: the number of equity activities initiated by the activity initiator within a preset time range and the total amount of equity, the number of equity activities initiated by the operator who initiates the equity activities within a preset time range and the total amount of equity, the initiation time of the equity activity, the start and end time range of the equity activity, the type of equity product, the total inventory of equity products, the restricted number of equity products per person, the clarity of product description, the failure rate of previous awarding of equity products provided by the same supplier, whether the activity involves jumping to a third-party link, whether the equity activity restricts participating customers, and at least one of the number of restricted participating customers in the equity activity.

9. A device for identifying suspicious equity activities, It is characterized in that include: A rights and interests activity audit information acquisition unit, used to acquire rights and interests activity audit information of a target rights and interests activity, wherein the target rights and interests activity is initiated for audit by an activity initiator in a rights and interests activity audit system and audited through a rights and interests activity audit process set in the rights and interests activity audit system, and the rights and interests activity audit information is obtained from the rights and interests activity audit system; A rights and interests activity and initiator information acquisition unit, used to acquire rights and interests activity information and activity initiator information of a target rights and interests activity; A feature extraction unit, configured to extract rights and interests activity review feature information from the rights and interests activity review information, and to extract activity and initiator feature information from the rights and interests activity information and the activity initiator information; An identification unit is used to determine the identification result of the target equity activity based on the equity activity review feature information, the activity and initiator feature information and a trained suspicious equity activity identification model, wherein the suspicious equity activity identification model is generated by generating training samples based on the equity activity review information, equity activity information and activity initiator information of multiple equity activities that have been reviewed and approved by the equity activity review system in the past, and then training a preset machine learning model based on the training samples.

10. The suspicious equity activity identification device according to claim 9, It is characterized in that Also includes: A historical data acquisition unit, used to acquire equity activity review information, equity activity information and activity initiator information of multiple equity activities that have been reviewed and approved by the equity activity review system in the past, and extract equity activity review feature information for model training and activity and initiator feature information for model training; A training sample generating unit, used to generate model input data for model training according to the equity activity review feature information used for model training and the activity and initiator feature information used for model training, and then set the recognition result label corresponding to the model input data used for model training to obtain a training sample; The model training unit is used to train the preset machine learning model according to the training samples to obtain the suspicious equity activity identification model.

11. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, It is characterized in that When the processor executes the computer program, the steps of the method according to any one of claims 1 to 8 are implemented.

12. A computer-readable storage medium having a computer program / instructions stored thereon, It is characterized in that When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

13. A computer program product comprising a computer program / instructions, It is characterized in that When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.