Behavior data evaluation method and device based on artificial intelligence, equipment and medium

Abstract

The invention belongs to the technical field of artificial intelligence, relates to a behavior data evaluation method and device based on artificial intelligence, computer equipment and a storage medium, is applied to the field of financial science and technology, and comprises the following steps: judging whether an insurance application request triggered by a user is received or not; if yes, extracting insurance buying behavior data from the insurance buying application request, and performing matching processing on the insurance buying behavior data based on a black product list library to obtain a matching result; if the matching result is that matching fails, preprocessing the insurance buying behavior data to obtain target insurance buying behavior data; performing prediction processing on the target insuring behavior data based on an insuring black production identification model to obtain a black production behavior prediction result; and generating a risk assessment result corresponding to the insurance buying behavior data based on the black production behavior prediction result. In addition, the risk assessment result can be stored in the block chain. According to the invention, based on the use of the insuring black production identification model, the processing efficiency and the processing accuracy of insuring black production risk identification can be effectively improved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence development technology and financial technology, and in particular to an artificial intelligence-based behavioral data evaluation method, device, computer equipment and storage medium. Background Art

[0002] In the insurance industry, illegal insurance activities have become common in recent years and have become a major risk faced by insurance companies and consumers. The illegal insurance activities refer to the behavior of some lawless elements who use illegal means, such as stealing and purchasing insurance policy information of insurance consumers, impersonating insurance company or regulatory department staff, and using fraudulent means to induce consumers to make abnormal cancellations or purchase new insurance products, and profit from it. This behavior not only seriously disrupts the normal order of the insurance market, but also greatly infringes on the legitimate rights and interests of insurance consumers, and also brings huge economic losses and reputation risks to life insurance companies.

[0003] Traditional methods of identifying risks of insurance black market mainly rely on personal experience and judgment. Risk assessors of insurance companies try to identify potential insurance black market behaviors by manually analyzing insurance information, customer behavior, etc. However, this manual assessment method requires a lot of manpower and cost. With the development of the insurance industry, the amount of insurance information has increased dramatically, and the amount of data that risk assessors need to process has also increased, resulting in low work efficiency and difficulty in coping with large-scale data analysis needs. In addition, the accuracy of manual assessment is low. Risk assessors are limited by their personal experience and knowledge level, and it is often difficult to accurately identify complex and changeable insurance black market behaviors. Summary of the invention

[0004] The purpose of the embodiments of the present application is to propose a behavioral data evaluation method, device, computer equipment and storage medium based on artificial intelligence to solve the technical problems of low processing efficiency and low accuracy in the existing methods for identifying risks of black industries underwriting based on manual evaluation.

[0005] In order to solve the above technical problems, the present application embodiment provides a behavior data evaluation method based on artificial intelligence, which adopts the following technical solutions:

[0006] Determine whether an insurance application request triggered by a user is received; wherein the insurance identification request carries insurance behavior data;

[0007] If yes, extract the insurance behavior data from the insurance application request, and match the insurance behavior data based on a preset black market list database to obtain a corresponding matching result;

[0008] Determine whether the matching result is a matching failure;

[0009] If the matching result is a match failure, preprocessing the insurance behavior data to obtain corresponding target insurance behavior data;

[0010] Based on the pre-trained insurance black industry identification model, the target insurance behavior data is predicted and processed to obtain the corresponding black industry behavior prediction results;

[0011] A risk assessment result corresponding to the insurance behavior data is generated based on the black industry behavior prediction result.

[0012] Furthermore, before the step of predicting the target insurance behavior data based on the pre-trained insurance black industry identification model to obtain the corresponding black industry behavior prediction result, it also includes:

[0013] Obtain pre-collected multi-dimensional data related to insurance;

[0014] Constructing feature data corresponding to the detection of the black market behavior of insurance based on the multi-dimensional data, and preprocessing the feature data to obtain target feature data;

[0015] Performing labeling processing on the target feature data to obtain labeled sample data;

[0016] Based on the sample data, a preset initial model is trained using a plurality of supervised learning methods to obtain a plurality of corresponding classification models;

[0017] Building a fusion model based on all the classification models;

[0018] Based on the confidence of the prediction result output by the fusion model, high-confidence sample data with a confidence level higher than a preset confidence threshold are screened out from the sample data;

[0019] generating target sample data based on the high confidence sample data and the sample data;

[0020] Based on the target sample data, the designated initial model is self-trained until a target model that meets the preset model effect indicators is obtained, and the target model is used as the insured black industry identification model.

[0021] Furthermore, the step of constructing a fusion model based on all the classification models specifically includes:

[0022] Get multiple preset fusion frameworks;

[0023] Filtering a target fusion frame from all the fusion frames;

[0024] Performing model building processing on all the classification models based on the target fusion framework to obtain a built processing model;

[0025] The processed model is used as the fusion model.

[0026] Furthermore, the step of preprocessing the insurance behavior data to obtain corresponding target insurance behavior data specifically includes:

[0027] Performing data cleaning processing on the insurance behavior data to obtain corresponding first data;

[0028] Performing standardization processing on the first data to obtain corresponding second data;

[0029] Performing data conversion processing on the second data to obtain corresponding third data;

[0030] Performing feature extraction processing on the third data to obtain corresponding fourth data;

[0031] The fourth data is used as the target insurance behavior data.

[0032] Furthermore, the step of generating a risk assessment result corresponding to the insurance behavior data based on the black market behavior prediction result specifically includes:

[0033] Obtaining a preset probability threshold;

[0034] Determine whether the black market behavior prediction result is greater than the probability threshold;

[0035] If so, generating a first risk assessment result indicating that the insurance behavior data has a risk of illegal activities;

[0036] If not, a second risk assessment result is generated indicating that the insurance behavior data does not contain any black industry risk.

[0037] Furthermore, before the step of matching the insurance behavior data based on the preset black market list database to obtain the corresponding matching result, it also includes:

[0038] Obtain recorded black market activities;

[0039] Obtaining the behavioral characteristics and behavioral patterns of the black market behavior;

[0040] Call the preset initial database;

[0041] The behavior characteristics and behavior patterns of the black market behavior are stored in the initial database to obtain the black market list database.

[0042] Furthermore, after the step of generating a risk assessment result corresponding to the insurance behavior data based on the black market behavior prediction result, the method further includes:

[0043] If the risk assessment result is that the insurance behavior data has a risk of illegal activities, then the response to the insurance application request is restricted;

[0044] Generate corresponding warning information based on the insurance behavior data and the risk assessment result;

[0045] Get the preset alarm notification method;

[0046] Based on the alarm notification method, the early warning information is sent to relevant personnel.

[0047] In order to solve the above technical problems, the embodiment of the present application further provides a behavior data evaluation device based on artificial intelligence, which adopts the following technical solution:

[0048] The first judgment module is used to judge whether an insurance application request triggered by a user is received; wherein the insurance identification request carries insurance behavior data;

[0049] A matching module, for extracting the insurance behavior data from the insurance application request, and performing matching processing on the insurance behavior data based on a preset black market list library to obtain a corresponding matching result;

[0050] A second judgment module is used to judge whether the matching result is a matching failure;

[0051] A preprocessing module, for preprocessing the insurance behavior data to obtain corresponding target insurance behavior data if the matching result is a matching failure;

[0052] A prediction module is used to predict the target insurance behavior data based on the pre-trained insurance black industry identification model to obtain the corresponding black industry behavior prediction result;

[0053] The first generating module is used to generate a risk assessment result corresponding to the insurance behavior data based on the black industry behavior prediction result.

[0054] 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:

[0055] Determine whether an insurance application request triggered by a user is received; wherein the insurance identification request carries insurance behavior data;

[0056] If yes, extract the insurance behavior data from the insurance application request, and match the insurance behavior data based on a preset black market list database to obtain a corresponding matching result;

[0057] Determine whether the matching result is a matching failure;

[0058] If the matching result is a match failure, preprocessing the insurance behavior data to obtain corresponding target insurance behavior data;

[0059] Based on the pre-trained insurance black industry identification model, the target insurance behavior data is predicted and processed to obtain the corresponding black industry behavior prediction results;

[0060] A risk assessment result corresponding to the insurance behavior data is generated based on the black industry behavior prediction result.

[0061] 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:

[0062] Determine whether an insurance application request triggered by a user is received; wherein the insurance identification request carries insurance behavior data;

[0063] If yes, extract the insurance behavior data from the insurance application request, and match the insurance behavior data based on a preset black market list database to obtain a corresponding matching result;

[0064] Determine whether the matching result is a matching failure;

[0065] If the matching result is a match failure, preprocessing the insurance behavior data to obtain corresponding target insurance behavior data;

[0066] Based on the pre-trained insurance black industry identification model, the target insurance behavior data is predicted and processed to obtain the corresponding black industry behavior prediction results;

[0067] A risk assessment result corresponding to the insurance behavior data is generated based on the black industry behavior prediction result.

[0068] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0069] The present application first determines whether an insurance application request triggered by a user is received; wherein the insurance identification request carries insurance behavior data; if so, the insurance behavior data is extracted from the insurance application request, and the insurance behavior data is matched based on a preset black market list library to obtain a corresponding matching result; then it is determined whether the matching result is a matching failure; if the matching result is a matching failure, the insurance behavior data is pre-processed to obtain the corresponding target insurance behavior data; subsequently, the target insurance behavior data is predicted based on a pre-trained insurance black market identification model to obtain the corresponding black market behavior prediction result; finally, a risk assessment result corresponding to the insurance behavior data is generated based on the black market behavior prediction result. Different from the existing method of identifying the insured black industry risks based on manual assessment, the present application is based on the use of a black industry list database and an insured black industry identification model, which can realize fast and accurate predictive processing of target insured behavior data to obtain black industry behavior prediction results, and then can generate risk assessment results corresponding to the insured behavior data based on the automatically generated black industry behavior prediction results, effectively improving the processing efficiency and accuracy of identifying the insured black industry risks of target insured behaviors, and improving the accuracy of the generated black industry behavior prediction results and risk assessment results. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] 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.

[0071] Figure 1 is an exemplary system architecture diagram to which the present application may be applied;

[0072] Figure 2 A flowchart of an embodiment of an artificial intelligence-based behavior data evaluation method according to the present application;

[0073] Figure 3 is a structural schematic diagram of an embodiment of an artificial intelligence-based behavior data evaluation device according to the present application;

[0074] Figure 4 It is a structural diagram of an embodiment of a computer device according to the present application. DETAILED DESCRIPTION

[0075] 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.

[0076] 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.

[0077] 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.

[0078] 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, etc.

[0079] 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.

[0080] The terminal device 101 can be any electronic device with a display screen and supporting web browsing. In addition to the laptop computer 1011, tablet computer 1012 or 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), a laptop computer, a desktop computer, etc.

[0081] 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 .

[0082] It should be noted that the artificial intelligence-based behavior data evaluation method provided in the embodiment of the present application is generally executed by a server / terminal device, and accordingly, the artificial intelligence-based behavior data evaluation device is generally arranged in the server / terminal device.

[0083] 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 implementation requirements.

[0084] Continue to refer Figure 2 , showing a flow chart of an embodiment of an artificial intelligence-based behavior data evaluation method according to the present application. According to different needs, the order of the steps in the flowchart can be changed, and some steps can be omitted. The artificial intelligence-based behavior data evaluation method provided in the embodiment of the present application can be applied to any scenario where black industry behavior identification is required, and the artificial intelligence-based behavior data evaluation method can be applied to products in these scenarios, for example, black industry behavior identification in the financial insurance field. The artificial intelligence-based behavior data evaluation method comprises the following steps:

[0085] Step S201, determining whether an insurance application request triggered by a user is received; wherein the insurance identification request carries insurance behavior data.

[0086] In this embodiment, the electronic device (e.g. Figure 1The server / terminal device shown in the figure) can obtain the insurance application request through a wired connection or a wireless connection. It should be pointed out that the above-mentioned wireless connection method may include but is not limited to 3G / 4G / 5G connection, WiFi connection, Bluetooth connection, WiMAX connection, Zigbee connection, UWB (ultra wideband) connection, and other wireless connection methods currently known or developed in the future. The executor of this application may be an intelligent underwriting system, or simply referred to as the system. The above-mentioned users may be customers who need to perform insurance application operations. The system can receive insurance behavior data input by the user through the front end or data source in real time through an API interface or a message queue. Among them, the insurance behavior data may include the customer's basic information, policy information, claims records, customer behavior, agent behavior, product information and other data.

[0087] Step S202: If yes, extract the insurance behavior data from the insurance application request, and match the insurance behavior data based on a preset black market list to obtain a corresponding matching result.

[0088] In this embodiment, the insurance application request can be parsed to extract the insurance behavior data contained in the insurance application request. Among them, the obtained insurance behavior data can be matched with the records in the above-mentioned black market list library. If there is a specified record in the black market list library that matches the insurance behavior data, a matching result of successful matching is generated, and it is determined that the insurance behavior data belongs to black market behavior, and the interception mechanism will be automatically triggered to refuse to respond to the insurance application request triggered by the user. If there is no specified record in the black market list library that matches the insurance behavior data, a matching result of failed matching is generated. In addition, for the construction process of the above-mentioned black market list library, this application will further describe this in detail in the subsequent specific embodiments, and will not be elaborated on here.

[0089] Step S203, determining whether the matching result is a matching failure.

[0090] In this embodiment, the content of the above matching result includes a successful match or a failed match.

[0091] Step S204: if the matching result is a matching failure, the insurance behavior data is pre-processed to obtain corresponding target insurance behavior data.

[0092] In this embodiment, the above-mentioned specific implementation process of preprocessing the insurance behavior data to obtain the corresponding target insurance behavior data will be further described in detail in subsequent specific embodiments of the present application, and will not be elaborated on here.

[0093] Step S205, predicting and processing the target insurance behavior data based on the pre-trained insurance black industry identification model to obtain corresponding black industry behavior prediction results.

[0094] In this embodiment, the target insurance behavior data can be input into a pre-trained insurance black industry identification model, and then the target insurance behavior data can be predicted and processed by the insurance black industry identification model, thereby outputting a black industry behavior prediction result corresponding to the target insurance behavior data. Among them, the above-mentioned black industry behavior prediction result is a prediction probability, specifically a probability of predicting the insurance behavior data as black industry behavior. In addition, the construction process of the above-mentioned insurance black industry identification model will be further described in detail in the subsequent specific embodiments of this application, and will not be elaborated on here.

[0095] Step S206, generating a risk assessment result corresponding to the insurance behavior data based on the black industry behavior prediction result.

[0096] In this embodiment, the specific implementation process of generating the risk assessment results corresponding to the insured behavior data based on the black market behavior prediction results will be described in further detail in subsequent specific embodiments of the present application and will not be elaborated on here.

[0097] The present application first determines whether an insurance application request triggered by a user is received; wherein the insurance identification request carries insurance behavior data; if so, the insurance behavior data is extracted from the insurance application request, and the insurance behavior data is matched based on a preset black market list library to obtain a corresponding matching result; then it is determined whether the matching result is a matching failure; if the matching result is a matching failure, the insurance behavior data is pre-processed to obtain the corresponding target insurance behavior data; subsequently, the target insurance behavior data is predicted based on a pre-trained insurance black market identification model to obtain the corresponding black market behavior prediction result; finally, a risk assessment result corresponding to the insurance behavior data is generated based on the black market behavior prediction result. Different from the existing method of identifying the insured black industry risks based on manual assessment, the present application is based on the use of a black industry list database and an insured black industry identification model, which can realize fast and accurate predictive processing of target insured behavior data to obtain black industry behavior prediction results, and then can generate risk assessment results corresponding to the insured behavior data based on the automatically generated black industry behavior prediction results, effectively improving the processing efficiency and accuracy of identifying the insured black industry risks of target insured behaviors, and improving the accuracy of the generated black industry behavior prediction results and risk assessment results.

[0098] In some optional implementations, before step S205, the electronic device may further perform the following steps:

[0099] Obtain pre-collected multi-dimensional data related to insurance.

[0100] In this embodiment, the multi-dimensional data may at least include basic customer information (such as age, gender, occupation, etc.), policy information (such as insurance type, insured amount, insurance time, etc.), claim records (such as number of claims, claim amount, etc.), customer behavior (such as login frequency, operating habits, etc.), agent behavior (such as agent sales records, customer feedback, etc.), product information (such as product popularity, complaint rate, etc.), etc. The collected multi-dimensional data can be carefully sorted and classified to ensure the integrity and consistency of the data, providing a solid foundation for subsequent model training.

[0101] Based on the multi-dimensional data, feature data corresponding to the detection of illegal insurance behaviors is constructed, and the feature data is pre-processed to obtain target feature data.

[0102] In this embodiment, a series of characteristic factors for detection can be designed and constructed according to the characteristics of the insurance black industry. These characteristic factors include but are not limited to frequent changes in insurance information, abnormally high insurance amounts, multiple insurances in a short period of time, lack of consistent insurance information, abnormal behavior of agents, abnormal claims behavior, abnormal customer behavior patterns, etc. In addition, the characteristic data can be processed by processing missing values, correcting erroneous data, removing duplicate data, etc., to eliminate noise and outliers in the characteristic data, thereby improving the data quality of the generated target characteristic data.

[0103] The target feature data is labeled to obtain labeled sample data.

[0104] In this embodiment, the target feature data can be labeled by machine labeling or manual labeling to obtain sample data of whether it is a normal sample or a black market sample.

[0105] Based on the sample data, a plurality of supervised learning methods are used to train the preset initial model to obtain corresponding multiple classification models.

[0106] In this embodiment, the supervised learning methods may include logistic regression (LR), support vector machine (SVM), random forest (RandomForest), gradient boosting decision tree (GBDT), LightGBM, multi-layer perceptron (MLP), etc. By using sample data and adopting various supervised learning methods to train the initial model, multiple classification models corresponding to the number of supervised learning methods that can preliminarily distinguish normal samples from black market samples are obtained.

[0107] A fusion model is constructed based on all the classification models.

[0108] In this embodiment, the specific implementation process of constructing the fusion model based on all the classification models will be further described in detail in subsequent specific embodiments of the present application and will not be elaborated in detail here.

[0109] Based on the confidence of the prediction result output by the fusion model, high-confidence sample data with a confidence threshold higher than a preset value is screened out from the sample data.

[0110] In this embodiment, the training results of multiple classification models are pieced together into a new feature matrix, and the meta-learner is input for training, evaluation and iteration to output the confidence of the prediction results of the fusion model. Among them, the role of the meta-learner is to further integrate the prediction results of multiple classification models to improve the accuracy and stability of the prediction. The prediction results of the fusion model are then evaluated for confidence to ensure the reliability of the prediction results. Confidence evaluation can be performed by calculating the prediction probability, setting the confidence threshold, etc. High-confidence prediction results are more likely to represent the true classification situation. In addition, according to the prediction results of the fusion model and actual needs, a reasonable confidence threshold (such as 0.7) is set. This confidence threshold is used to screen high-confidence self-annotated data (i.e., high-confidence sample data). Self-annotated data below the confidence threshold are subsequently discarded because the prediction results of these self-annotated data below the confidence threshold may not be reliable enough. And accept high-confidence sample data and output it as a normal sample or a black production sample. These samples will be used for subsequent iterative training of the insured black production identification model.

[0111] Target sample data is generated based on the high confidence sample data and the sample data.

[0112] In this embodiment, new target sample data is generated by adding the selected high-confidence sample data to the above sample data, which helps to expand the scale of sample data and improve the generalization ability of the model.

[0113] Based on the target sample data, the designated initial model is self-trained until a target model that meets the preset model effect indicators is obtained, and the target model is used as the insured black industry identification model.

[0114] In this embodiment, the designated initial model is iteratively processed by the Self-Training model by using the target sample data containing the labeled and self-generated labeled data. Specifically, the designated initial model is first retrained using the target sample data. The training process at this stage is similar to the classification model training stage, but the target sample data used has been updated and expanded. Furthermore, the labeled data set (i.e., high-confidence sample data) is used to perform model evaluation, including indicators such as AUC, KS, recall rate, and coverage rate. According to the evaluation results, the performance of the designated initial model is optimized, and the model parameters and algorithm strategies are adjusted. Repeat the above model training steps until the preset model effect indicators or the maximum number of iterations are met to obtain the final target model and use it as an insured black industry identification model to continuously improve the accuracy and robustness of insured black industry identification.

[0115] The present application obtains pre-collected multi-dimensional data related to insurance; and constructs feature data corresponding to the detection of insurance black market behavior based on the multi-dimensional data, and pre-processes the feature data to obtain target feature data; then annotates the target feature data to obtain annotated sample data; and based on the sample data, uses a variety of supervised learning methods to train the preset initial model to obtain corresponding multiple classification models; then constructs a fusion model based on all the classification models; and based on the confidence of the prediction result output by the fusion model, screens out high-confidence sample data above the preset confidence threshold from the sample data; subsequently generates target sample data based on the high-confidence sample data and the sample data; finally, based on the target sample data, performs self-training on the specified initial model until a target model that meets the preset model effect index is obtained, and the target model is used as the insurance black market identification model. This application combines data preprocessing, model training, model fusion and confidence assessment, automated data labeling, and self-training model iteration of multi-dimensional data related to insurance, and through continuous optimization and improvement, can build an efficient and accurate insurance black industry identification model, thereby improving the construction efficiency of the insurance black industry identification model and ensuring the model effect of the generated insurance black industry identification model.

[0116] In some optional implementations of this embodiment, the building of a fusion model based on all the classification models includes the following steps:

[0117] Get multiple preset fusion frameworks.

[0118] In this embodiment, the fusion framework may include at least a single-layer fusion framework or a multi-layer fusion framework.

[0119] A target fusion frame is selected from all the fusion frames.

[0120] In this embodiment, a framework may be randomly selected from the multiple fusion frameworks as the target fusion framework.

[0121] Based on the target fusion framework, model building processing is performed on all the classification models to obtain a built processing model.

[0122] In this embodiment, the training results of multiple classification models are pieced together into a new feature matrix by using a target fusion framework, and a corresponding fusion model is built, which helps to combine the advantages of different classification models and improve the overall prediction performance.

[0123] The processed model is used as the fusion model.

[0124] In this embodiment, compared with a single classification model, the generated fusion model has the advantages of stronger generalization ability, better robustness, and higher diversity, which is beneficial to reducing the risk of labeling insurance black market result data.

[0125] This application obtains a plurality of preset fusion frameworks; then selects a target fusion framework from all the fusion frameworks; then performs model building processing on all the classification models based on the target fusion framework to obtain a constructed processing model; and subsequently uses the processing model as the fusion model. This application selects a target fusion framework from the plurality of fusion frameworks obtained, and then performs model building processing on all the classification models based on the target fusion framework, so that the corresponding fusion model can be constructed quickly and accurately. Since the generated fusion model has the advantages of stronger generalization ability, better robustness, and higher diversity, it is beneficial to reduce the risk of sample annotation corresponding to the insured black industry result data.

[0126] In some optional implementations, step S203 includes the following steps:

[0127] The insurance behavior data is cleaned to obtain corresponding first data.

[0128] In this embodiment, the data cleaning process includes processing of missing values, correcting erroneous data, and removing duplicate data.

[0129] The first data is standardized to obtain corresponding second data.

[0130] In this embodiment, the above-mentioned standardization processing includes converting the data into a distribution with a mean of 0 and a standard deviation of 1, so as to eliminate the influence of different dimensions on the data.

[0131] Perform data conversion processing on the second data to obtain corresponding third data.

[0132] In this embodiment, the second data may be converted according to business needs, such as converting categorical variables into numerical variables (using one-hot encoding, label encoding, etc.) to obtain corresponding third data.

[0133] Perform feature extraction processing on the third data to obtain corresponding fourth data.

[0134] In this embodiment, the data type of the fourth data that has been processed by feature extraction is the same as the data type of the feature data corresponding to the detection of the insured black market behavior constructed based on the multi-dimensional data.

[0135] The fourth data is used as the target insurance behavior data.

[0136] The present application performs data cleaning on the insurance behavior data to obtain the corresponding first data; then performs standardization on the first data to obtain the corresponding second data; then performs data conversion on the second data to obtain the corresponding third data; subsequently performs feature extraction on the third data to obtain the corresponding fourth data; and finally uses the fourth data as the target insurance behavior data. The present application performs data cleaning, standardization, data conversion, and feature extraction on the insurance behavior data, thereby achieving fast and intelligent preprocessing of the insurance behavior data, and effectively improving the data quality and consistency of the generated target insurance behavior data.

[0137] In some optional implementations, step S206 includes the following steps:

[0138] Get the preset probability threshold.

[0139] In this embodiment, there is no specific limitation on the value of the above probability threshold, which can be set according to actual business needs, for example, it can be set to 0.75. If the probability of the prediction of the black industry behavior output by the insured black industry identification model exceeds the probability threshold, it is considered that the corresponding insured behavior has the risk of black industry.

[0140] Determine whether the black market behavior prediction result is greater than the probability threshold.

[0141] In this embodiment, the above-mentioned black industry behavior prediction result can specifically be a probability of predicting the insurance behavior data as black industry behavior. By numerically comparing the above-mentioned black industry behavior prediction result with the above-mentioned probability threshold, a corresponding numerical comparison result is obtained. Among them, the numerical comparison result includes that the black industry behavior prediction result is greater than the probability threshold, or the black industry behavior prediction result is less than the probability threshold.

[0142] If so, a first risk assessment result is generated indicating that the insurance behavior data has a risk of illegal activities.

[0143] In this embodiment, if the black industry behavior prediction result is greater than the probability threshold, it is determined that the insurance behavior data has a black industry risk, and a first risk assessment result indicating that the insurance behavior data has a black industry risk is generated.

[0144] If not, a second risk assessment result is generated indicating that the insurance behavior data does not contain any black industry risk.

[0145] In this embodiment, if the black industry behavior prediction result is less than the probability threshold, it is determined that the insurance behavior data does not have a black industry risk, and a first risk assessment result that the insurance behavior data does not have a black industry risk is generated.

[0146] This application obtains a preset probability threshold; then determines whether the black industry behavior prediction result is greater than the probability threshold; if so, generates a first risk assessment result that the insured behavior data has a black industry risk; if not, generates a second risk assessment result that the insured behavior data does not have a black industry risk. This application obtains a preset probability threshold, then numerically compares the above-mentioned black industry behavior prediction result with the above-mentioned probability threshold, and then can accurately generate the corresponding risk assessment result of the insured behavior data based on the obtained numerical comparison result, thereby improving the intelligence of the risk assessment result generation and ensuring the accuracy of the obtained risk assessment result.

[0147] In some optional implementations of this embodiment, before step S202, the electronic device may further perform the following steps:

[0148] Get the recorded black market activities.

[0149] In this embodiment, during the business processing of black market behavior identification, the detected black market behavior will be stored and a unique identifier will be set for the black market behavior.

[0150] Obtain the behavioral characteristics and behavioral patterns of the black market behavior.

[0151] In this embodiment, while recording the illegal activities, the behavior characteristics and behavior patterns of the illegal activities are also stored.

[0152] Call the preset initial database.

[0153] In this embodiment, the initial database may be a blank database table or file.

[0154] The behavior characteristics and behavior patterns of the black market behavior are stored in the initial database to obtain the black market list database.

[0155] In this embodiment, the recorded behavior characteristics and behavior patterns of the black market behavior are stored in the above-mentioned initial database to obtain a corresponding black market list library containing the unique identification, behavior characteristics, behavior patterns and other information of the black market behavior.

[0156] In addition, the black market list database can be regularly maintained and optimized, outdated or invalid records can be deleted, and newly collected black market behavior characteristics and behavior patterns can be updated. At the same time, the black market list database can be continuously improved and optimized based on actual application conditions and feedback data to improve the accuracy of identification and interception.

[0157] This application obtains the recorded black industry behavior; then obtains the behavioral characteristics and behavioral patterns of the black industry behavior; then calls the preset initial database; and subsequently stores the behavioral characteristics and behavioral patterns of the black industry behavior in the initial database to obtain the black industry list library. This application obtains the behavioral characteristics and behavioral patterns of the recorded black industry behavior, and then stores the behavioral characteristics and behavioral patterns of the black industry behavior in the initial database, so that the corresponding black industry list library can be quickly and intelligently constructed, thereby improving the construction efficiency and intelligence of the black industry list library. It is beneficial to use the black industry list library for rapid identification and interception in the subsequent insurance behavior review process. And when the new insurance behavior matches the record in the black industry list library, it can automatically trigger the interception mechanism or perform manual review.

[0158] In some optional implementations of this embodiment, after step S206, the electronic device may further perform the following steps:

[0159] If the risk assessment result is that the insurance behavior data contains black market risks, the response processing to the insurance application request is restricted.

[0160] In this embodiment, if the risk assessment result is that the insurance behavior data contains black market risks, the insurance application corresponding to the insurance behavior data will be rejected.

[0161] Corresponding warning information is generated based on the insurance behavior data and the risk assessment results.

[0162] In this embodiment, the insurance behavior data and the risk assessment results can be processed based on the preset warning rules to generate corresponding warning information. The warning information may include key information such as insurance information (such as the name of the insured, ID number, insured product, etc.), risk assessment results (such as the probability of black market behavior, predicted category, etc.) and risk level (such as high risk, medium risk, low risk, etc.).

[0163] Get the preset alarm notification method.

[0164] In this embodiment, there is no specific limitation on the selection of the above-mentioned alarm notification method, and for example, it may include email, text message, instant messaging tool, system notification and the like.

[0165] Based on the alarm notification method, the early warning information is sent to relevant personnel.

[0166] In this embodiment, the generated warning information can be promptly and accurately conveyed to relevant personnel through the above-mentioned alarm notification method.

[0167] If the present application detects that the risk assessment result is that the insured behavior data has a black industry risk, the response processing to the insurance application request will be restricted; then a corresponding warning information will be generated based on the insured behavior data and the risk assessment result; then a preset alarm notification method will be obtained; and subsequently, based on the alarm notification method, the warning information will be sent to relevant personnel. If the present application detects that the risk assessment result is that the insured behavior data has a black industry risk, the response processing to the insurance application request will be intelligently restricted, and corresponding warning information will be generated based on the insured behavior data and the risk assessment result, and then the warning information will be sent to relevant personnel in a timely and accurate manner based on the alarm notification method, so that after the relevant personnel receive the warning information, they can take corresponding response measures according to the warning content, thereby helping to improve the work experience of the relevant personnel, and improve the risk prevention and control of insured behaviors with black industry risks.

[0168] In some optional implementations, the user information obtained is subject to the user's consent and complies with relevant laws and policies.

[0169] In addition, the non-Company software tools or components that appear in the embodiments of the present application are merely examples and do not represent actual use.

[0170] In addition, the present application provides a method for identifying the insured black industry based on self-training (Self-Training) + model fusion technology. This method uses automatic data annotation and self-training + model fusion to develop a semi-supervised (Semi-supervised Learning, SSL) model, which can identify illegal and fraudulent behaviors involving the insurance field and effectively respond to the threat of the insured black industry. By combining the active learning and iterative optimization capabilities of the Self-Training technology and the model fusion method to improve the confidence of the labeled data, this method can effectively reduce the cost of data annotation, gradually improve the accuracy and robustness of the model, better respond to the variation of the insured black industry, and provide a more reliable insured black industry identification solution.

[0171] The insured black industry identification method of this application expenditure can generate the following value:

[0172] 1. Reduce the risk of insurance black market: Based on this application, it can intelligently identify potential illegal and fraudulent behaviors, help insurance companies to promptly discover and prevent insurance black market activities, improve the security and reliability of insurance business, protect the interests of insurance companies and consumers, and enhance the sustainable development of insurance business.

[0173] 2. Reduce the cost of black market labeling: This application uses only a small amount of labeled data to automatically predict and label unlabeled data, reducing the overall cost of manual labeling and improving the efficiency and accuracy of data labeling.

[0174] 3. Improve the confidence of labeled samples: Introduce model fusion to conduct confidence assessment on labeled data. On the basis of effectively improving the accuracy of labeled data, minimize the risk of labeling results and ensure the development effect of the insured black industry identification model.

[0175] 4. Improve the adaptability of black market methods: This application can effectively adapt to the ever-changing black market methods and control insurance risks by continuously adding new black market data to the iterative upgrade process of the self-learning model.

[0176] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0177] It should be emphasized that in order to further ensure the privacy and security of the above risk assessment results, the above risk assessment results can also be stored in a blockchain node.

[0178] The blockchain referred to in this application is a new application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanism, encryption algorithm, etc. Blockchain is essentially a decentralized database, a string of data blocks generated by cryptographic methods. Each data block contains a batch of network transaction information, which is used to verify the validity of its information (anti-counterfeiting) and generate the next block. Blockchain can include the underlying blockchain platform, platform product service layer, and application service layer.

[0179] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, 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.

[0180] 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.

[0181] 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).

[0182] 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.

[0183] Further references Figure 3 , as a response to the above Figure 2 The present application provides an embodiment of a behavior data evaluation device based on artificial intelligence. Figure 2 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.

[0184] like Figure 3 As shown, the artificial intelligence-based behavior data evaluation device 300 described in this embodiment includes: a first judgment module 301, a matching module 302, a second judgment module 303, a preprocessing module 304, a prediction module 305 and a first generation module 306. Among them:

[0185] The first judgment module 301 is used to judge whether an insurance application request triggered by a user is received; wherein the insurance identification request carries insurance behavior data;

[0186] The matching module 302 is used to extract the insurance behavior data from the insurance application request, and perform matching processing on the insurance behavior data based on a preset black market list library to obtain a corresponding matching result;

[0187] The second judgment module 303 is used to judge whether the matching result is a matching failure;

[0188] A preprocessing module 304 is used to preprocess the insurance behavior data to obtain corresponding target insurance behavior data if the matching result is a matching failure;

[0189] Prediction module 305, used to perform prediction processing on the target insurance behavior data based on the pre-trained insurance black industry identification model to obtain the corresponding black industry behavior prediction result;

[0190] The first generating module 306 is used to generate a risk assessment result corresponding to the insurance behavior data based on the black industry behavior prediction result.

[0191] In this embodiment, the operations performed by the above modules or units respectively correspond one-to-one to the steps of the artificial intelligence-based behavior data evaluation method in the aforementioned embodiment, and will not be repeated here.

[0192] In some optional implementations of this embodiment, the behavior data evaluation device based on artificial intelligence further includes:

[0193] A first acquisition module is used to acquire pre-collected multi-dimensional data related to insurance;

[0194] A first processing module is used to construct feature data corresponding to the detection of insurance black market behavior based on the multi-dimensional data, and pre-process the feature data to obtain target feature data;

[0195] A module annotation block is used to perform annotation processing on the target feature data to obtain annotated sample data;

[0196] A training module, used to train a preset initial model based on the sample data using a variety of supervised learning methods to obtain a corresponding plurality of classification models;

[0197] A construction module, used for constructing a fusion model based on all the classification models;

[0198] A screening module, used for screening out high-confidence sample data higher than a preset confidence threshold from the sample data based on the confidence of the prediction result output by the fusion model;

[0199] A second generating module, configured to generate target sample data based on the high confidence sample data and the sample data;

[0200] The second processing module is used to perform self-training processing on the designated initial model based on the target sample data until a target model that meets the preset model effect indicators is obtained, and the target model is used as the insured black industry identification model.

[0201] In this embodiment, the operations performed by the above modules or units respectively correspond one-to-one to the steps of the artificial intelligence-based behavior data evaluation method in the aforementioned embodiment, and will not be repeated here.

[0202] In some optional implementations of this embodiment, the building blocks include:

[0203] The first acquisition submodule is used to acquire a plurality of preset fusion frames;

[0204] A screening submodule, used for screening out a target fusion frame from all the fusion frames;

[0205] A building submodule is used to perform model building processing on all the classification models based on the target fusion framework to obtain a built processing model;

[0206] The first determining submodule is used to use the processing model as the fusion model.

[0207] In this embodiment, the operations performed by the above modules or units respectively correspond one-to-one to the steps of the artificial intelligence-based behavior data evaluation method in the aforementioned embodiment, and will not be repeated here.

[0208] In some optional implementations of this embodiment, the preprocessing module 304 includes:

[0209] A first processing submodule is used to perform data cleaning processing on the insurance behavior data to obtain corresponding first data;

[0210] A second processing submodule, used for performing standardization processing on the first data to obtain corresponding second data;

[0211] A third processing submodule, used for performing data conversion processing on the second data to obtain corresponding third data;

[0212] A fourth processing submodule, used for performing feature extraction processing on the third data to obtain corresponding fourth data;

[0213] The second determining submodule is used to use the fourth data as the target insurance behavior data.

[0214] In this embodiment, the operations performed by the above modules or units respectively correspond one-to-one to the steps of the artificial intelligence-based behavior data evaluation method in the aforementioned embodiment, and will not be repeated here.

[0215] In some optional implementations of this embodiment, the first generating module 306 includes:

[0216] The second acquisition submodule is used to obtain a preset probability threshold;

[0217] A judgment submodule, used to judge whether the prediction result of the black market behavior is greater than the probability threshold;

[0218] A first generating submodule is used to generate a first risk assessment result that the insurance behavior data has a black industry risk if yes;

[0219] The second generating submodule is used to generate a second risk assessment result indicating that the insurance behavior data does not have any black industry risk.

[0220] In this embodiment, the operations performed by the above modules or units respectively correspond one-to-one to the steps of the artificial intelligence-based behavior data evaluation method in the aforementioned embodiment, and will not be repeated here.

[0221] In some optional implementations of this embodiment, the behavior data evaluation device based on artificial intelligence further includes:

[0222] The second acquisition module is used to obtain recorded black market activities;

[0223] The third acquisition module is used to obtain the behavior characteristics and behavior patterns of the black market behavior;

[0224] A calling module is used to call a preset initial database;

[0225] The storage module is used to store the behavior characteristics and behavior patterns of the black market behavior in the initial database to obtain the black market list library.

[0226] In this embodiment, the operations performed by the above modules or units respectively correspond one-to-one to the steps of the artificial intelligence-based behavior data evaluation method in the aforementioned embodiment, and will not be repeated here.

[0227] In some optional implementations of this embodiment, the behavior data evaluation device based on artificial intelligence further includes:

[0228] A third processing module is used to limit the response processing of the insurance application request if the risk assessment result is that the insurance behavior data has a black industry risk;

[0229] A third generating module, configured to generate corresponding warning information based on the insurance behavior data and the risk assessment result;

[0230] The fourth acquisition module is used to obtain a preset alarm notification method;

[0231] The sending module is used to send the warning information to relevant personnel based on the alarm notification method.

[0232] In this embodiment, the operations performed by the above modules or units respectively correspond one-to-one to the steps of the artificial intelligence-based behavior data evaluation method in the aforementioned embodiment, and will not be repeated here.

[0233] To solve the above technical problems, the present application also provides a computer device. Figure 4 , Figure 4 This is a basic structural block diagram of the computer device in this embodiment.

[0234] The computer device 4 includes a memory 41, a processor 42, and a network interface 43 that are interconnected through a system bus. It should be noted that the figure only shows a computer device 4 with components 41-43, but it should be understood that it is not required to implement all the components shown, and more or fewer components can 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 microprocessors, application specific integrated circuits (Application Specific Integrated Circuit, ASIC), programmable gate arrays (Field-Programmable Gate Array, FPGA), digital processors (Digital Signal Processor, DSP), embedded devices, etc.

[0235] The computer device may be a desktop computer, a notebook, a PDA, a cloud server, etc. The computer device may interact with the user through a keyboard, a mouse, a remote controller, a touch pad, or a voice control device.

[0236] The memory 41 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, etc. In some embodiments, the memory 41 can be an internal storage unit of the computer device 4, such as a hard disk or memory of the computer device 4. In other embodiments, the memory 41 can also be an external storage device of the computer device 4, such as a plug-in hard disk equipped on the computer device 4, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (FlashCard), etc. Of course, the memory 41 can also include both the internal storage unit of the computer device 4 and its external storage device. In this embodiment, the memory 41 is generally used to store the operating system and various application software installed on the computer device 4, such as computer-readable instructions of the behavior data evaluation method based on artificial intelligence. In addition, the memory 41 can also be used to temporarily store various types of data that have been output or are to be output.

[0237] The processor 42 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chip in some embodiments. The processor 42 is generally used to control the overall operation of the computer device 4. In this embodiment, the processor 42 is used to run computer-readable instructions or process data stored in the memory 41, such as computer-readable instructions for running the artificial intelligence-based behavior data evaluation method.

[0238] The network interface 43 may include a wireless network interface or a wired network interface. The network interface 43 is generally used to establish a communication connection between the computer device 4 and other electronic devices.

[0239] The present application also provides another embodiment, 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 at least one processor to enable the at least one processor to perform the steps of the above-mentioned artificial intelligence-based behavioral data evaluation method.

[0240] 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.

[0241] 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 behavioral data evaluation method based on artificial intelligence, characterized in that: The steps include: Determine whether an insurance application request triggered by a user is received; wherein the insurance identification request carries insurance behavior data; If yes, extract the insurance behavior data from the insurance application request, and match the insurance behavior data based on a preset black market list database to obtain a corresponding matching result; Determine whether the matching result is a matching failure; If the matching result is a match failure, preprocessing the insurance behavior data to obtain corresponding target insurance behavior data; Based on the pre-trained insurance black industry identification model, the target insurance behavior data is predicted and processed to obtain the corresponding black industry behavior prediction results; A risk assessment result corresponding to the insurance behavior data is generated based on the black industry behavior prediction result.

2. The artificial intelligence-based behavior data evaluation method according to claim 1, characterized in that: Before the step of predicting and processing the target insurance behavior data based on the pre-trained insurance black industry identification model to obtain the corresponding black industry behavior prediction result, the step further includes: Obtain pre-collected multi-dimensional data related to insurance; Constructing feature data corresponding to the detection of the black market behavior of insurance based on the multi-dimensional data, and preprocessing the feature data to obtain target feature data; Performing labeling processing on the target feature data to obtain labeled sample data; Based on the sample data, a preset initial model is trained using a plurality of supervised learning methods to obtain a plurality of corresponding classification models; Building a fusion model based on all the classification models; Based on the confidence of the prediction result output by the fusion model, high-confidence sample data with a confidence level higher than a preset confidence threshold are screened out from the sample data; generating target sample data based on the high confidence sample data and the sample data; Based on the target sample data, the designated initial model is self-trained until a target model that meets the preset model effect indicators is obtained, and the target model is used as the insured black industry identification model.

3. The artificial intelligence-based behavior data evaluation method according to claim 2, characterized in that: The step of building a fusion model based on all the classification models specifically includes: Get multiple preset fusion frameworks; Filtering a target fusion frame from all the fusion frames; Performing model building processing on all the classification models based on the target fusion framework to obtain a built processing model; The processed model is used as the fusion model.

4. The artificial intelligence-based behavior data evaluation method according to claim 1, characterized in that: The step of preprocessing the insurance behavior data to obtain corresponding target insurance behavior data specifically includes: Performing data cleaning processing on the insurance behavior data to obtain corresponding first data; Performing standardization processing on the first data to obtain corresponding second data; Performing data conversion processing on the second data to obtain corresponding third data; Performing feature extraction processing on the third data to obtain corresponding fourth data; The fourth data is used as the target insurance behavior data.

5. The artificial intelligence-based behavior data evaluation method according to claim 1, characterized in that: The step of generating a risk assessment result corresponding to the insurance behavior data based on the black industry behavior prediction result specifically includes: Obtaining a preset probability threshold; Determine whether the black market behavior prediction result is greater than the probability threshold; If so, generating a first risk assessment result indicating that the insurance behavior data has a risk of illegal activities; If not, a second risk assessment result is generated indicating that the insurance behavior data does not contain any black industry risk.

6. The artificial intelligence-based behavior data evaluation method according to claim 1, characterized in that: Before the step of matching the insurance behavior data based on the preset black market list database to obtain the corresponding matching result, the method further includes: Obtain recorded black market activities; Obtaining the behavioral characteristics and behavioral patterns of the black market behavior; Call the preset initial database; The behavior characteristics and behavior patterns of the black market behavior are stored in the initial database to obtain the black market list database.

7. The artificial intelligence-based behavior data evaluation method according to claim 1, characterized in that: After the step of generating a risk assessment result corresponding to the insurance behavior data based on the black market behavior prediction result, the method further includes: If the risk assessment result is that the insurance behavior data has a risk of illegal activities, then the response to the insurance application request is restricted; Generate corresponding warning information based on the insurance behavior data and the risk assessment result; Get the preset alarm notification method; Based on the alarm notification method, the early warning information is sent to relevant personnel.

8. A behavior data evaluation device based on artificial intelligence, characterized in that: include: The first judgment module is used to judge whether an insurance application request triggered by a user is received; wherein the insurance identification request carries insurance behavior data; A matching module, for extracting the insurance behavior data from the insurance application request, and performing matching processing on the insurance behavior data based on a preset black market list library to obtain a corresponding matching result; A second judgment module is used to judge whether the matching result is a matching failure; A preprocessing module, for preprocessing the insurance behavior data to obtain corresponding target insurance behavior data if the matching result is a matching failure; A prediction module is used to predict the target insurance behavior data based on the pre-trained insurance black industry identification model to obtain the corresponding black industry behavior prediction result; The first generating module is used to generate a risk assessment result corresponding to the insurance behavior data based on the black industry behavior prediction result.

9. A computer device, characterized in that: It comprises a memory and a processor, wherein the memory stores computer-readable instructions, and when the processor executes the computer-readable instructions, the steps of the artificial intelligence-based behavior data evaluation method as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the artificial intelligence-based behavior data evaluation method as described in any one of claims 1 to 7.

Images