CN115809930A - Anti-fraud analysis method, device, equipment and medium based on data fusion matching - Google Patents
Anti-fraud analysis method, device, equipment and medium based on data fusion matching Download PDFInfo
- Publication number
- CN115809930A CN115809930A CN202211591669.3A CN202211591669A CN115809930A CN 115809930 A CN115809930 A CN 115809930A CN 202211591669 A CN202211591669 A CN 202211591669A CN 115809930 A CN115809930 A CN 115809930A
- Authority
- CN
- China
- Prior art keywords
- data
- fusion
- fraud
- settlement
- probability
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Abstract
The invention relates to a data fusion technology, and discloses an anti-fraud analysis method based on data fusion matching, which comprises the following steps: performing data fusion on the basic information data, the credit data and the claim settlement data to obtain historical fusion data; constructing a risk decision tree according to the historical fusion data, and inputting the fusion data corresponding to the user into the risk decision tree to obtain a risk score value of the user; determining a fusion weight of the fusion data, and calculating the claim settlement probability of the user according to the risk score value and the fusion weight; inputting the claim probability and the claim data in the fusion data into an anti-fraud scheme matching model to obtain an anti-fraud scheme corresponding to the user; and determining the claim amount of the user according to a claim compensation mode in the anti-fraud scheme, and calculating the feasibility of claim settlement according to the claim probability, the claim data and the claim amount. The invention also provides an anti-fraud analysis device, electronic equipment and a storage medium based on data fusion matching. The invention can improve the feasibility of anti-fraud analysis.
Description
Technical Field
The invention relates to the technical field of data fusion, in particular to an anti-fraud analysis method and device based on data fusion matching, electronic equipment and a computer-readable storage medium.
Background
With the increasingly fierce competition of insurance market, cost reduction and efficiency improvement are urgently needed for a company management layer, in order to improve the overall operation level, analyze and summarize insurance risk reasons, rationality of a claim settlement process, obtain risk rules, reduce life and property losses of people and enterprise claim costs, analysis of the claim settlement process needs to be carried out according to various data behaviors of users, and the users are guaranteed not to be cheated in the claim settlement process.
The existing claim settlement analysis technology is mostly completed by manual operation based on the claim settlement operation flow, and each claim is audited and rechecked manually. In practical application, the method is completely completed by manual operation, and the risk of the user cannot be strictly controlled only based on the claim material submitted by the user, and the claim processing time is too long, so that the user is only considered to obtain single data, the risk of the user cannot be known, the claim processing time is too long, and the feasibility in anti-fraud analysis is low.
Disclosure of Invention
The invention provides an anti-fraud analysis method and device based on data fusion matching and a computer readable storage medium, and mainly aims to solve the problem of low feasibility in anti-fraud analysis.
In order to achieve the above object, the present invention provides an anti-fraud analysis method based on data fusion matching, which includes:
s1, acquiring basic information data, credit data and claim settlement data of a historical user, and performing data fusion on the basic information data, the credit data and the claim settlement data to obtain historical fusion data;
s2, constructing a risk decision tree according to the historical fusion data, and inputting fusion data corresponding to a target user into the risk decision tree to obtain a risk score value of the target user;
s3, determining a fusion weight of the fusion data by using a preset analytic hierarchy process, and calculating a claim settlement probability of the target user according to the risk score and the fusion weight by using a preset dissimilarity weight algorithm, wherein the claim settlement probability of the target user is calculated according to the risk score and the fusion weight by using the preset dissimilarity weight algorithm, and the method comprises the following steps:
s31, acquiring the amount of the claim of the insurance type purchased by the target;
s32, calculating the claim settlement probability of the target user according to the claim amount, the risk score value and the fusion weight by using a dissimilarity weight algorithm as follows:
wherein P is the claim probability, x k The amount of the claim of the kth insurance category, r the number of the insurance categories, alpha the fusion weight and tau the risk score value;
s4, inputting the claim probability and the claim data in the fused data into a pre-constructed anti-fraud scheme matching model to obtain an anti-fraud scheme corresponding to the target user;
and S5, determining the claim amount of the target user according to a claim compensation mode in the anti-fraud scheme, and calculating the feasibility of the anti-fraud scheme according to the claim probability, the claim data and the claim amount.
Optionally, the performing data fusion on the basic information data, the credit data, and the claim settlement data to obtain history fused data includes:
performing source division on the basic information data, the credit data and the claim settlement data according to preset data source requirements to obtain multi-source data;
performing dimensionality division on the basic information data, the credit data and the claim settlement data according to a preset data dimensionality requirement to obtain multidimensional data;
and performing data fusion on the multi-source data and the multi-dimensional data to obtain the historical fusion data.
Optionally, the performing data fusion on the multi-source data and the multi-dimensional data to obtain the historical fusion data includes:
extracting a first data characteristic vector of the multi-source data through a preset bag-of-words model, and extracting a second data characteristic vector of the multi-dimensional data through a preset text vector extraction model;
calculating the similarity of the first data feature vector and the second data feature vector by using a weighted similarity formula as follows:
wherein S is the similarity, delta i Is the ith data feature weight, K i Is the ith vector feature in the first data feature vector, H i The ith vector feature in the second data feature vector is obtained, and n is the number of the data vectors;
and performing data fusion on the multi-source data and the multi-dimensional data according to the similarity to obtain the historical fusion data.
Optionally, the constructing a risk decision tree according to the history fusion data includes:
extracting a characteristic data set in the historical fusion data;
calculating the information entropy of each feature data in the feature data set by using the following information entropy formula:
wherein G is the information entropy, p j Log is a logarithmic function which is the proportion of j-th class characteristic data, C is the total number of characteristic data samples in the characteristic data set, C j The number of characteristic attribute samples in the j-th type characteristic data is obtained;
selecting the characteristic data with the maximum information entropy as a root node, and splitting a left node and a right node on the root node;
selecting the feature data with the maximum information entropy from the unselected feature data set one by one, and distributing the feature data to the left node and the right node to obtain a basic decision tree;
and performing decision tree addition processing on the basic decision tree to obtain a risk decision tree.
Optionally, the inputting the fusion data corresponding to the target user into the risk decision tree to obtain the risk score value of the target user includes:
matching the first data characteristic of the fusion data with the data characteristic in the risk decision tree to obtain a risk level;
and determining the risk score value of the target user according to the risk grade.
Optionally, the inputting the claim probability and the claim data in the fused data into a pre-constructed anti-fraud scheme matching model to obtain an anti-fraud scheme corresponding to the target user includes:
inputting the claim settlement probability into the anti-fraud scheme matching model to obtain an anti-fraud classification scheme;
matching the claim data with the claim data in the anti-fraud classification scheme to obtain a matching value;
and selecting the anti-fraud classification scheme with the maximum matching value as the anti-fraud scheme corresponding to the target user.
Optionally, the calculating feasibility of the anti-fraud scheme according to the claim settlement probability, the claim settlement data and the claim settlement amount includes:
determining the claim weight of the claim data by using a preset entropy algorithm;
determining a first feasibility value of the claim amount according to the claim weight;
determining a second feasibility value of the claim amount according to the claim settlement probability;
determining the feasibility of the anti-fraud scheme based on an average of the first feasibility value and the second feasibility value.
In order to solve the above problem, the present invention further provides an anti-fraud analysis apparatus based on data fusion matching, the apparatus including:
the data fusion module is used for acquiring basic information data, credit data and claim settlement data of a historical user, and performing data fusion on the basic information data, the credit data and the claim settlement data to obtain historical fusion data;
the risk score value determining module is used for constructing a risk decision tree according to the historical fusion data and inputting the fusion data corresponding to the target user into the risk decision tree to obtain the risk score value of the target user;
the claim settlement probability calculation module is used for determining a fusion weight of the fusion data by using a preset analytic hierarchy process, and calculating the claim settlement probability of the target user according to the risk score value and the fusion weight through a preset dissimilarity weight algorithm;
the anti-fraud scheme determining module is used for inputting the claim probability and the claim data in the fused data into a pre-constructed anti-fraud scheme matching model to obtain an anti-fraud scheme corresponding to the target user;
and the anti-fraud scheme feasibility calculating module is used for determining the claim amount of the target user according to the claim compensation mode in the anti-fraud scheme and calculating the feasibility of the anti-fraud scheme according to the claim probability, the claim data and the claim amount.
In order to solve the above problem, the present invention also provides an electronic device, including:
at least one processor; and the number of the first and second groups,
a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,
the memory stores a computer program executable by the at least one processor, the computer program being executable by the at least one processor to enable the at least one processor to perform the above-described data fusion matching-based anti-fraud analysis method.
In order to solve the above problem, the present invention further provides a computer-readable storage medium, in which at least one computer program is stored, and the at least one computer program is executed by a processor in an electronic device to implement the above-mentioned anti-fraud analysis method based on data fusion matching.
According to the embodiment of the invention, the historical data of multiple aspects of the user is acquired, the historical data is subjected to data fusion, and valuable comprehensive information which cannot be obtained by single or single source data can be obtained based on the fused data. And then constructing a risk decision tree according to the fusion data, and when risk prediction is performed on the user, inputting the fusion data of the user into the risk decision tree so as to obtain a risk score value of the user. And then calculating the claim settlement probability of the user according to the risk score value and the fusion weight of the fusion data, and judging whether the user is subjected to claim settlement or not according to the claim settlement probability, wherein the risk score value of the user is too high, and the claim settlement probability is lower. The claim probability and the claim data in the fusion data are input into an anti-fraud scheme matching model, an optimal anti-fraud scheme can be obtained, the claim amount is further determined according to a claim compensation mode in the anti-fraud scheme, and the feasibility of the claim can be analyzed according to the claim probability, the claim data and the claim amount, so that the user and an insurance company are satisfied, the user is guaranteed not to be cheated in the claim settlement process, and the feasibility of the anti-fraud scheme is improved. Therefore, the anti-fraud analysis method, the anti-fraud analysis device, the electronic equipment and the computer-readable storage medium based on data fusion matching can solve the problem of low feasibility during anti-fraud analysis.
Drawings
Fig. 1 is a schematic flow chart of an anti-fraud analysis method based on data fusion matching according to an embodiment of the present invention;
FIG. 2 is a flow chart illustrating fusion data according to an embodiment of the present invention;
FIG. 3 is a flowchart illustrating a process of calculating a probability of claim settlement according to an embodiment of the present invention;
FIG. 4 is a functional block diagram of an anti-fraud analysis apparatus based on data fusion matching according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of an electronic device for implementing the anti-fraud analysis method based on data fusion matching according to an embodiment of the present invention.
The implementation, functional features and advantages of the present invention will be further described with reference to the accompanying drawings.
Detailed Description
It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
The embodiment of the application provides an anti-fraud analysis method based on data fusion matching. The execution subject of the anti-fraud analysis method based on data fusion matching includes, but is not limited to, at least one of electronic devices such as a server and a terminal that can be configured to execute the method provided by the embodiments of the present application. In other words, the anti-fraud analysis method based on data fusion matching may be executed by software or hardware installed in the terminal device or the server device, and the software may be a block chain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, and the like. The server may be an independent server, or may be a cloud server that provides basic cloud computing services such as a cloud service, a cloud database, cloud computing, a cloud function, cloud storage, a Network service, cloud communication, a middleware service, a domain name service, a security service, a Content Delivery Network (CDN), a big data and artificial intelligence platform, and the like.
Fig. 1 is a schematic flow chart of an anti-fraud analysis method based on data fusion matching according to an embodiment of the present invention. In this embodiment, the anti-fraud analysis method based on data fusion matching includes:
s1, acquiring basic information data, credit data and claim settlement data of a historical user, and performing data fusion on the basic information data, the credit data and the claim settlement data to obtain historical fusion data;
in one practical application scenario of the present invention, claims are important matters about benefits of insurance companies, different customers have completely different requirements, and different profitability is provided for the insurance companies, but potentially high-risk customers may exist among many customers, and claims for the high-risk customers directly cause claim expenses of the insurance companies. Therefore, it is necessary to merge data of customers on multiple platforms and analyze the risk level of the user according to the merged data.
In the embodiment of the invention, the basic information data comprises the age, the sex, the occupation identity, the income information, the asset information, the health condition and the like of the user; the credit data comprises transaction records, loan information, repayment information and the like of the user; the claim data comprises a claim insurance policy, a claim beneficiary, a claim reason and the like.
In detail, the stored basic information data, credit data and claim data can be fetched from a predetermined storage area by using a computer sentence (such as java sentence, python sentence, etc.) with a data fetching function, wherein the storage area comprises but is not limited to a database, a block chain node and a network cache.
In the embodiment of the invention, different data are fused and associated. For example, when a user is subject to claim settlement processing, the basic information data of the user is checked, whether the user is a high-risk user is checked according to the credit data of the user, and when the user is not a high-risk user, the claim settlement data of the user is checked, so that a claim is settled on the user.
In the embodiment of the present invention, referring to fig. 2, the performing data fusion on the basic information data, the credit data, and the claim settlement data to obtain history fusion data includes:
s21, performing source division on the basic information data, the credit data and the claim settlement data according to preset data source requirements to obtain multi-source data;
s22, performing dimensionality division on the basic information data, the credit data and the claim settlement data according to a preset data dimensionality requirement to obtain multi-dimensional data;
and S23, performing data fusion on the multi-source data and the multi-dimensional data to obtain historical fusion data.
In detail, the data source requirement mainly refers to a main source of data, such as a Web end, a transaction platform, a bank, an insurance platform, and the like. The data dimension requirements mainly refer to various attributes of data, for example, the dimensions of basic information data comprise the age, sex, occupational identity, income information, asset information, health condition and the like of a user; the dimension of the credit data comprises transaction records, loan information, repayment information and the like; the dimensions of the claims data include the claims policy, the beneficiary of the claim, the reason for the claim, and the like. For example, the age may be from a plurality of data platforms, and different types of data may contain the same attribute, such as the basic information data is divided according to the dimension requirement to obtain a plurality of attributes, namely, the age, the gender, the professional identity, the income information, and the like.
Specifically, the multi-source data is divided according to the source of the data, the dividing standard of the multi-dimensional data is divided according to the attribute of the data, the multi-source data can also be regarded as one dimension of the multi-dimension, and the meaning of the multi-dimensional data is higher than that of the multi-source data. However, there is no absolute relationship between multi-source data and multi-dimensional data, data of a single source can be divided into multiple dimensions according to different properties, and data of the same property can be divided into multiple sources according to different sources. Therefore, the multi-source data and the multi-dimensional data are processing objects of data fusion, wherein the data fusion is attribute fusion and is to combine data from multiple sources and process the data by using a computer to obtain valuable comprehensive information which cannot be obtained by a single source or a single type of source.
In the embodiment of the present invention, the performing data fusion on the multi-source data and the multi-dimensional data to obtain the historical fusion data includes:
extracting a first data characteristic vector of the multi-source data through a preset bag-of-word model, and extracting a second data characteristic vector of the multi-dimensional data through a preset text vector extraction model;
calculating the similarity of the first data feature vector and the second data feature vector by using a weighted similarity formula as follows:
wherein S is the similarity, delta i Is the ith data feature weight, K i Is the ith vector feature in the first data feature vector, H i The ith vector feature in the second data feature vector is obtained, and n is the number of the data vectors;
and performing data fusion on the multi-source data and the multi-dimensional data according to the similarity to obtain the historical fusion data.
In detail, the first data feature is to perform feature extraction on the multi-source feature through a preset word bag model, that is, firstly, performing Chinese word segmentation on multi-source data, storing all the appeared words into one table, and then obtaining vector representation of each segmented word according to whether each word appears in the word list, wherein the word appearing in the word list is marked as 1, and the word not appearing in the word list is marked as 0, so that the vector representation of each multi-source data or multi-dimensional data is obtained. Wherein each dimension of the vector uniquely corresponds to a word in the vocabulary. The text vector extraction model is a model for generating word vectors based on the word2vec model.
Specifically, the data weighted feature weight δ in the weighted similarity formula can be used to represent the importance of data in multi-source data and multi-dimensional data, and further judge the similarity between different data features according to the importance of the data, the greater the weighted feature weight between the data is, the more important the data is represented, the data with the same importance are clustered together, which is beneficial to the fusion between the data, and the comprehensive analysis of the risk value of the user is improved.
Further, data between the multi-source data and the multi-dimensional data are clustered together according to the similarity, namely when the similarity between the data is larger than a preset similarity threshold value, the data are clustered in a set, so that fusion between the data is realized, and valuable comprehensive information which cannot be obtained by a single source or a single type of source can be obtained from multiple aspects according to the data fusion.
Furthermore, when the multi-source data and the multi-dimensional data are subjected to data fusion, whether the user is a high-risk user or not can be evaluated from the comprehensive data, so that the behavior of the user can be determined.
S2, constructing a risk decision tree according to the historical fusion data, and inputting fusion data corresponding to a target user into the risk decision tree to obtain a risk score value of the target user;
in the embodiment of the invention, the nodes in the risk decision tree are represented by the characteristics of the case for establishing the decision tree model, and the branches of the decision tree are represented by the values of the case characteristics. And the characteristic variables adopted by the risk decision tree analysis are the data characteristics of the historical fusion data, and the analysis is focused on predicting the risk level of the user.
In an embodiment of the present invention, the constructing a risk decision tree according to the history fusion data includes:
extracting a characteristic data set in the historical fusion data;
calculating the information entropy of each feature data in the feature data set by using the following information entropy formula:
wherein G is the information entropy, p j Log is a logarithmic function which is the proportion of j-th class characteristic data, C is the total number of characteristic data samples in the characteristic data set, C j The number of characteristic attribute samples in the j-th type characteristic data is obtained;
selecting the characteristic data with the maximum information entropy as a root node, and splitting a left node and a right node on the root node;
selecting the feature data with the maximum information entropy from the unselected feature data sets one by one, and distributing the feature data to the left node and the right node to obtain a basic decision tree;
and adding decision trees to the basic decision trees to obtain risk decision trees.
In detail, the feature data in the feature data set in the history fusion data includes an insurance amount, a user age, an affiliated occupation code, a claim amount, a credit score, and the like. Wherein, the attributes of the insurance amount are classified into low-level insurance, medium-level insurance, high-level insurance and the like; the age of the user is classified into young, middle-aged and old.
Specifically, the purity of each feature data can be determined according to the information entropy formula, whether the branch nodes belong to the same category can be determined, and the generated risk decision tree can be used for predicting the risk score of the user more accurately. And selecting the feature data with the maximum information entropy as a root node, selecting the feature data with the maximum information entropy from the unselected feature data on the basis of the root node, and adding the feature data into the left node or the right node of the root node until all the feature data are selected, indicating that the risk decision tree is aggregated, thereby obtaining the risk decision tree model.
Further, in the process of training the decision tree, in order to classify the training samples as correctly as possible, the node division process is repeated continuously, which sometimes causes too many branches of the decision tree, and at this time, some characteristics of the training set are considered as general properties of all data to cause overfitting due to the good learning of the training samples. Therefore, the risk of overfitting is reduced by actively adding processing to the decision tree.
In the embodiment of the invention, the risk decision tree is constructed to predict the risk of the user, and the next operation of the user is limited and judged according to the risk value.
In the embodiment of the present invention, the inputting the fusion data corresponding to the target user into the risk decision tree to obtain the risk score value of the target user includes:
matching the first data characteristics of the fusion data with the data characteristics in the risk decision tree to obtain a risk level;
and determining the risk score value of the target user according to the risk grade.
In detail, the first data characteristics of the fusion data comprise insurance amount, user age, belonging occupational code, claim amount, credit score and the like of the current user, and the risk grade of the final user can be obtained by matching the data characteristics in the pre-constructed risk decision tree step by step, wherein the risk grade is divided into low grade, medium grade and high grade.
Specifically, when the risk grade is low, the risk score value is 0 to 20; when the risk grade is a middle grade, the risk score value is 20-60; when the risk rating is high, the risk score value is 60-100. The higher the risk score value, the higher the risk that indicates the user.
Illustratively, when the claim aspect is health insurance, the age is most correlated with the risk level, and generally, the greater the risk, the higher the chance of hospitalization.
Furthermore, the claim settlement probability of the user is further analyzed according to the risk score value of the user, and whether the claim is settled on the user or not can be further judged according to the claim settlement probability.
S3, determining a fusion weight of the fusion data by using a preset analytic hierarchy process, and calculating the claim settlement probability of the target user according to the risk score value and the fusion weight through a preset dissimilarity weight algorithm;
in the embodiment of the present invention, the fusion weight is an expression for sorting importance of data features in the fusion data. The fusion weight of the fusion data may also be different for claims of different insurance. For example, in health insurance, the fusion weight of age characteristics is relatively high, while in car insurance, the fusion weight of car usage and car specific accidents is relatively high.
In detail, the analytic hierarchy process is a decision process for decomposing elements always related to decision into levels of targets, criteria, schemes and the like, and performing qualitative and quantitative analysis on the basis, and is a hierarchical weight decision analysis process. Firstly, a hierarchical structure model is established, a decision target, a considered factor (decision criterion) and a decision object are divided into a highest layer, a middle layer and a lowest layer according to the mutual relation among the decision target, the considered factor (decision criterion) and the decision object, and a hierarchical structure diagram is drawn. The highest level refers to the purpose of decision making and the problem to be solved; the lowest layer refers to an alternative scheme in decision making; the middle layer refers to the considered factors and decision criteria; for two adjacent layers, the upper layer is called a target layer, and the lower layer is called a factor layer. Determining a feature matrix of the fusion data; calculating a weight vector of the fusion data; and carrying out normalization processing on the weight vector to obtain the fusion weight of the fusion data.
In the embodiment of the invention, the claim settlement probability is to analyze the claim settlement events of the user, determine how much probability the user can carry out claim settlement, and when the claim settlement probability of the user is too low, the user can carry out no claim settlement or proper claim settlement; when the claim settlement probability of the user is too high, the user is subjected to claim settlement.
In an embodiment of the present invention, referring to fig. 3, the calculating, by a preset dissimilarity weight algorithm, a claim settlement probability of the target user according to the risk score and the fusion weight includes:
s31, acquiring the amount of the claim of the insurance type purchased by the target;
s32, calculating the claim settlement probability of the target user according to the claim amount, the risk score value and the fusion weight by using an diversification weight algorithm as follows:
wherein P is the claim probability, x k The sum of claims of the kth insurance category, r is the number of insurance categories, α is the fusion weight, and τ is the risk score value.
In detail, when an insured life suffers a loss due to the occurrence of a particular risk event within an insurance contract, the insurer is under the responsibility of paying the insurance premium. The amount of the claim amount varies with different types of insurance compensation modes, but does not exceed the actual loss value caused by the event. Such contingent payments linked to a loss value are called claim payments. It is obvious that the loss value or amount of the insurance target is a random variable, and the amount of the corresponding claim paid by the claim is determined by the estimation of the probability of the claim.
Specifically, the claim settlement probability is related to the claim amount, the fusion weight and the risk score value. The fusion weight represents the characteristic weight of the user in different insurance types, which influences the calculation of the claim probability, and the claim amount is a measure for calculating the claim probability of the user in the claim process.
Illustratively, when a user purchases four insurance types, the amount of money to be paid for each insurance is 1 ten thousand yuan, 2 ten thousand yuan, 3 ten thousand yuan and 4 ten thousand yuan, and the fusion weight of the user is 0.7, but the risk score value of the user is 90 points, the probability of claim settlement of the user calculated according to the formula is relatively low, and fraud is likely to exist, so that the user needs to be checked next to determine whether to settle the claim.
S4, inputting the claim settlement probability and the claim settlement data in the fusion data into a pre-constructed anti-fraud scheme matching model to obtain an anti-fraud scheme corresponding to the target user;
in the embodiment of the invention, the anti-fraud scheme matching model is obtained by utilizing a large number of anti-fraud schemes for training based on a random forest, wherein the random forest is a classifier composed of a set of decision trees, and each tree is constructed by applying an algorithm and adding a random vector on a training set. And analyzing the optimal anti-fraud scheme of the user according to the original insurance purchase scheme, the claim settlement accident information and the claim settlement probability of the user so as to determine the optimal claim settlement amount of the user and enable the user and an insurance company to achieve the best result.
In an embodiment of the present invention, the inputting the claim settlement probability and the claim settlement data in the fused data into a pre-constructed anti-fraud scheme matching model to obtain the anti-fraud scheme corresponding to the target user includes:
inputting the claim settlement probability into the anti-fraud scheme matching model to obtain an anti-fraud classification scheme;
matching the claim data with the claim data in the anti-fraud classification scheme to obtain a matching value;
and selecting the anti-fraud classification scheme with the maximum matching value as the anti-fraud scheme corresponding to the target user.
In detail, the anti-fraud scheme which accords with the claim probability can be classified by inputting the claim probability into the anti-fraud scheme matching model. For example, when the anti-fraud scheme matches the corresponding anti-fraud scheme with a probability of claims below fifty percent in the model as a, the corresponding anti-fraud scheme with a probability of claims above fifty percent is as B. The anti-fraud scheme corresponding to the claim probability is found in the anti-fraud scheme matching model.
Specifically, after the anti-fraud scheme corresponding to the claim probability is found in the anti-fraud scheme matching model, the anti-fraud scheme with the maximum matching degree is found according to the check matching between the claim insurance policy, the claim beneficiary, the claim reason and the like contained in the claim data submitted among users and the anti-fraud classification scheme.
Further, after determining the optimal anti-fraud scheme, the policy contract and other data of the insurance policy are further checked against each other, thereby determining the amount of the anti-fraud scheme and the manner of settlement.
And S5, determining the claim amount of the target user according to a claim compensation mode in the anti-fraud scheme, and calculating the feasibility of the anti-fraud scheme according to the claim probability, the claim data and the claim amount.
In the embodiment of the invention, the claim settlement compensation mode is to settle the claim for the target user according to different insurance types and insurance limits. Wherein the insurance type is injury risk, health risk, accident risk, etc. The insurance amount is based on the quota insurance originally owned by the insurance purchase.
In detail, when determining the anti-fraud scheme, the user may be settled according to the insurance quota, a part of the settlement may be performed on the user, the user may be rejected, and the settlement amount is different in different forms of anti-fraud schemes. Therefore, the user is required to be settled according to the specific claim data and the claim settlement conditions submitted by the user. For example, when a user takes an airplane and has an accident, an insurance company can carry out claims settlement on the user according to the insurance quota when the user purchases the accident; when the user drives a car and has an accident, part of the accident is the responsibility of the user, and the insurance company can carry out part of settlement; when the user makes fraud, forges insurance materials or hides information, the insurance company refuses payment.
Further, after the anti-fraud scheme and the claim amount are determined, the feasibility of the anti-fraud scheme is determined, and the user and the insurance company are guaranteed to achieve the best results.
In an embodiment of the present invention, the calculating the feasibility of the anti-fraud scheme according to the claim settlement probability, the claim settlement data, and the claim settlement amount includes:
determining the claim settlement weight of the claim data by using a preset entropy algorithm;
determining a first feasibility value of the claim amount according to the claim weight;
determining a second feasibility value of the claim amount according to the claim settlement probability;
determining the feasibility of the anti-fraud scheme based on the mean of the first feasibility value and the second feasibility value.
In detail, the claim weight is used to represent authenticity of the claim data, and an amount of claim that should be claim to the user according to the claim data is determined according to the authenticity of the claim data, so as to determine accuracy of the amount of claim, i.e. a first feasibility of implementation of the amount of claim, i.e. when the claim weight is proportional to the first feasibility value of the amount of claim, when the claim weight is 0.9, the first feasibility value of the amount of claim is 90, and when the claim weight is 0.2, the first feasibility value of the amount of claim is 20. As such. And when the claim settlement probability is 80%, the second feasibility value of the claim settlement amount is 80, and when the claim settlement probability is 30%, the second feasibility value of the claim settlement amount is 30.
Specifically, the feasibility of each anti-fraud scheme is determined according to the average value of the first feasibility value and the second feasibility value, that is, when the first feasibility value is 90 and the second feasibility value is 80, the feasibility of the claim is 85, that is, the claim is feasible for both the user and the insurance company; if the first feasibility value is 50 and the second feasibility value is 50, the feasibility of the claim settlement is 50, and the claim settlement may have disputes between the user and the insurance company, and further processing needs to be performed according to the feasibility, so that the user and the insurance company can obtain satisfactory results.
According to the embodiment of the invention, the historical data of multiple aspects of the user is acquired, the historical data is subjected to data fusion, and valuable comprehensive information which cannot be obtained by single or single source data can be obtained based on the fused data. And then constructing a risk decision tree according to the fusion data, and inputting the fusion data of the user into the risk decision tree when risk prediction is carried out on the user, so as to obtain a risk score value of the user. And then calculating the claim settlement probability of the user according to the risk score value and the fusion weight of the fusion data, and judging whether to claim the user according to the claim settlement probability, wherein the risk score value of the user is too high, and the claim settlement probability is lower. The claim probability and the claim data in the fused data are input into an anti-fraud scheme matching model, an optimal anti-fraud scheme can be obtained, the claim amount is further determined according to a claim compensation mode in the anti-fraud scheme, and the feasibility of the claim can be analyzed according to the claim probability, the claim data and the claim amount, so that the user and an insurance company are satisfied, the user is guaranteed not to be cheated in the claim settlement process, and the feasibility of the anti-fraud scheme is improved. Therefore, the anti-fraud analysis method, the anti-fraud analysis device, the anti-fraud analysis electronic equipment and the anti-fraud analysis computer-readable storage medium based on data fusion matching can solve the problem of low feasibility in claim settlement analysis.
Fig. 4 is a functional block diagram of an anti-fraud analysis apparatus based on data fusion matching according to an embodiment of the present invention.
The anti-fraud analysis apparatus 100 based on data fusion matching according to the present invention can be installed in an electronic device. According to the realized functions, the anti-fraud analysis device 100 based on data fusion matching can comprise a data fusion module 101, a risk score value determination module 102, a claim settlement probability calculation module 103, an anti-fraud scheme determination module 104 and an anti-fraud scheme feasibility calculation module 105. The module of the present invention, which may also be referred to as a unit, refers to a series of computer program segments that can be executed by a processor of an electronic device and that can perform a fixed function, and that are stored in a memory of the electronic device.
In the present embodiment, the functions regarding the respective modules/units are as follows:
the data fusion module 101 is configured to acquire basic information data, credit data, and claim settlement data of a historical user, and perform data fusion on the basic information data, the credit data, and the claim settlement data to obtain historical fusion data;
the risk score value determining module 102 is configured to construct a risk decision tree according to the historical fusion data, and input fusion data corresponding to a target user into the risk decision tree to obtain a risk score value of the target user;
the claim settlement probability calculation module 103 is configured to determine a fusion weight of the fusion data by using a preset analytic hierarchy process, and calculate a claim settlement probability of the target user according to the risk score and the fusion weight through a preset dissimilarity weight algorithm;
the anti-fraud scheme determining module 104 is configured to input the claim settlement probability and the claim settlement data in the fused data into a pre-constructed anti-fraud scheme matching model, so as to obtain an anti-fraud scheme corresponding to the target user;
the anti-fraud scheme feasibility calculating module 105 is configured to determine a claim amount of the target user according to a claim compensation manner in the anti-fraud scheme, and calculate the feasibility of the anti-fraud scheme according to the claim probability, the claim data, and the claim amount.
In detail, when the modules in the anti-fraud analysis apparatus 100 based on data fusion matching according to the embodiment of the present invention are used, the same technical means as the anti-fraud analysis method based on data fusion matching described in fig. 1 to fig. 3 are used, and the same technical effect can be produced, which is not described again here.
Fig. 5 is a schematic structural diagram of an electronic device for implementing an anti-fraud analysis method based on data fusion matching according to an embodiment of the present invention.
The electronic device 1 may comprise a processor 10, a memory 11, a communication bus 12 and a communication interface 13, and may further comprise a computer program, such as an anti-fraud analysis program based on data fusion matching, stored in the memory 11 and executable on the processor 10.
In some embodiments, the processor 10 may be composed of an integrated circuit, for example, a single packaged integrated circuit, or may be composed of a plurality of integrated circuits packaged with the same function or different functions, and includes one or more Central Processing Units (CPUs), a microprocessor, a digital Processing chip, a graphics processor, a combination of various control chips, and the like. The processor 10 is a Control Unit (Control Unit) of the electronic device, connects various components of the whole electronic device by using various interfaces and lines, and executes various functions and processes data of the electronic device by running or executing programs or modules stored in the memory 11 (for example, executing an anti-fraud analysis program based on data fusion matching, etc.), and calling data stored in the memory 11.
The memory 11 includes at least one type of readable storage medium including flash memory, removable hard disks, multimedia cards, card-type memory (e.g., SD or DX memory, etc.), magnetic memory, magnetic disks, optical disks, and the like. The memory 11 may in some embodiments be an internal storage unit of the electronic device, for example a removable hard disk of the electronic device. The memory 11 may also be an external storage device of the electronic device in other embodiments, such as a plug-in mobile hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), and the like, which are provided on the electronic device. Further, the memory 11 may also include both an internal storage unit and an external storage device of the electronic device. The memory 11 may be used not only to store application software installed in the electronic device and various types of data, such as codes of an anti-fraud analysis program based on data fusion matching, etc., but also to temporarily store data that has been output or is to be output.
The communication bus 12 may be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. The bus may be divided into an address bus, a data bus, a control bus, etc. The bus is arranged to enable connection communication between the memory 11 and at least one processor 10 or the like.
The communication interface 13 is used for communication between the electronic device and other devices, and includes a network interface and a user interface. Optionally, the network interface may include a wired interface and/or a wireless interface (e.g., WI-FI interface, bluetooth interface, etc.), which are typically used to establish a communication connection between the electronic device and other electronic devices. The user interface may be a Display (Display), an input unit such as a Keyboard (Keyboard), and optionally a standard wired interface, a wireless interface. Alternatively, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch device, or the like. The display, which may also be referred to as a display screen or display unit, is suitable, among other things, for displaying information processed in the electronic device and for displaying a visualized user interface.
Only electronic devices having components are shown, and those skilled in the art will appreciate that the structures shown in the figures do not constitute limitations on the electronic devices, and may include fewer or more components than shown, or some components in combination, or a different arrangement of components.
For example, although not shown, the electronic device may further include a power supply (such as a battery) for supplying power to each component, and preferably, the power supply may be logically connected to the at least one processor 10 through a power management device, so that functions of charge management, discharge management, power consumption management and the like are realized through the power management device. The power supply may also include any component of one or more dc or ac power sources, recharging devices, power failure detection circuitry, power converters or inverters, power status indicators, and the like. The electronic device may further include various sensors, a bluetooth module, a Wi-Fi module, and the like, which are not described herein again.
It is to be understood that the described embodiments are for purposes of illustration only and that the scope of the appended claims is not limited to such structures.
The anti-fraud analysis program based on data fusion matching stored in the memory 11 of the electronic device 1 is a combination of a plurality of instructions, which when executed in the processor 10, can realize:
acquiring basic information data, credit data and claim settlement data of a historical user, and performing data fusion on the basic information data, the credit data and the claim settlement data to obtain historical fusion data;
constructing a risk decision tree according to the historical fusion data, and inputting fusion data corresponding to a target user into the risk decision tree to obtain a risk score value of the target user;
determining a fusion weight of the fusion data by using a preset analytic hierarchy process, and calculating the claim settlement probability of the target user according to the risk score value and the fusion weight through a preset dissimilarity weight algorithm;
inputting the claim probability and the claim data in the fused data into a pre-constructed anti-fraud scheme matching model to obtain an anti-fraud scheme corresponding to the target user;
and determining the claim amount of the target user according to a claim compensation mode in the anti-fraud scheme, and calculating the feasibility of the anti-fraud scheme according to the claim probability, the claim data and the claim amount.
Specifically, the specific implementation method of the instruction by the processor 10 may refer to the description of the relevant steps in the embodiment corresponding to the drawings, which is not described herein again.
Further, the integrated modules/units of the electronic device 1, if implemented in the form of software functional units and sold or used as separate products, may be stored in a computer readable storage medium. The computer readable storage medium may be volatile or non-volatile. For example, the computer-readable medium may include: any entity or device capable of carrying said computer program code, a recording medium, a usb-disk, a removable hard disk, a magnetic diskette, an optical disk, a computer Memory, a Read-Only Memory (ROM).
The present invention also provides a computer-readable storage medium, storing a computer program which, when executed by a processor of an electronic device, may implement:
acquiring basic information data, credit data and claim settlement data of a historical user, and performing data fusion on the basic information data, the credit data and the claim settlement data to obtain historical fusion data;
constructing a risk decision tree according to the historical fusion data, and inputting fusion data corresponding to a target user into the risk decision tree to obtain a risk score value of the target user;
determining a fusion weight of the fusion data by using a preset analytic hierarchy process, and calculating the claim settlement probability of the target user according to the risk score value and the fusion weight through a preset dissimilarity weight algorithm;
inputting the claim probability and the claim data in the fused data into a pre-constructed anti-fraud scheme matching model to obtain an anti-fraud scheme corresponding to the target user;
and determining the claim amount of the target user according to a claim compensation mode in the anti-fraud scheme, and calculating the feasibility of the anti-fraud scheme according to the claim probability, the claim data and the claim amount.
In the embodiments provided in the present invention, it should be understood that the disclosed apparatus, device and method can be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the modules is only one logical functional division, and other divisions may be realized in practice.
The modules described as separate parts may or may not be physically separate, and parts displayed as modules may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, functional modules in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, or in a form of hardware plus a software functional module.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof.
The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference signs in the claims shall not be construed as limiting the claim concerned.
Furthermore, it is obvious that the word "comprising" does not exclude other elements or steps, and the singular does not exclude the plural. A plurality of units or means recited in the system claims may also be implemented by one unit or means in software or hardware. The terms first, second, etc. are used to denote names, but not any particular order.
Finally, it should be noted that the above embodiments are only intended to illustrate the technical solutions of the present invention and not to limit the same, and although the present invention is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.
Claims (10)
1. An anti-fraud analysis method based on data fusion matching, characterized in that the method comprises:
s1, acquiring basic information data, credit data and claim settlement data of a historical user, and performing data fusion on the basic information data, the credit data and the claim settlement data to obtain historical fusion data;
s2, constructing a risk decision tree according to the historical fusion data, and inputting fusion data corresponding to a target user into the risk decision tree to obtain a risk score value of the target user;
s3, determining a fusion weight of the fusion data by using a preset analytic hierarchy process, and calculating a claim settlement probability of the target user according to the risk score and the fusion weight by using a preset dissimilarity weight algorithm, wherein the claim settlement probability of the target user is calculated according to the risk score and the fusion weight by using the preset dissimilarity weight algorithm, and the method comprises the following steps:
s31, acquiring the amount of the claim of the insurance type purchased by the target;
s32, calculating the claim settlement probability of the target user according to the claim amount, the risk score value and the fusion weight by using an diversification weight algorithm as follows:
wherein P is the claim probability, x k The amount of the claim of the kth insurance category, r the number of the insurance categories, alpha the fusion weight and tau the risk score value;
s4, inputting the claim settlement probability and the claim settlement data in the fusion data into a pre-constructed anti-fraud scheme matching model to obtain an anti-fraud scheme corresponding to the target user;
and S5, determining the claim amount of the target user according to a claim compensation mode in the anti-fraud scheme, and calculating the feasibility of the anti-fraud scheme according to the claim probability, the claim data and the claim amount.
2. The anti-fraud analysis method based on data fusion matching according to claim 1, wherein the data fusion of the basic information data, the credit data and the claim settlement data to obtain historical fusion data includes:
according to preset data source requirements, source division is carried out on the basic information data, the credit data and the claim settlement data to obtain multi-source data;
performing dimensionality division on the basic information data, the credit data and the claim settlement data according to a preset data dimensionality requirement to obtain multidimensional data;
and performing data fusion on the multi-source data and the multi-dimensional data to obtain the historical fusion data.
3. The anti-fraud analysis method based on data fusion matching according to claim 2, wherein the data fusion of the multi-source data and the multi-dimensional data to obtain the historical fusion data comprises:
extracting a first data characteristic vector of the multi-source data through a preset bag-of-word model, and extracting a second data characteristic vector of the multi-dimensional data through a preset text vector extraction model;
calculating the similarity of the first data feature vector and the second data feature vector by using a weighted similarity formula as follows:
wherein S is the similarity, delta i Is the ith data feature weight, K i Is the ith vector feature in the first data feature vector, H i The ith vector feature in the second data feature vector is obtained, and n is the number of the data vectors;
and performing data fusion on the multi-source data and the multi-dimensional data according to the similarity to obtain the historical fusion data.
4. The method for anti-fraud analysis based on data fusion matching according to claim 1, wherein the constructing a risk decision tree based on the historical fusion data comprises:
extracting a characteristic data set in the historical fusion data;
calculating the information entropy of each feature data in the feature data set by using the following information entropy formula:
wherein G is the information entropy, p j Log is a logarithmic function which is the proportion of j-th class characteristic data, C is the total number of characteristic data samples in the characteristic data set, C j The number of characteristic attribute samples in the j-th type characteristic data is set;
selecting the characteristic data with the maximum information entropy as a root node, and splitting a left node and a right node on the root node;
selecting the feature data with the maximum information entropy from the unselected feature data sets one by one, and distributing the feature data to the left node and the right node to obtain a basic decision tree;
and performing decision tree addition processing on the basic decision tree to obtain a risk decision tree.
5. The method for anti-fraud analysis based on data fusion matching according to claim 1, wherein the inputting the fusion data corresponding to the target user into the risk decision tree to obtain the risk score value of the target user includes:
matching the first data characteristic of the fusion data with the data characteristic in the risk decision tree to obtain a risk level;
and determining the risk score value of the target user according to the risk grade.
6. The anti-fraud analysis method based on data fusion matching according to any one of claims 1 to 5, wherein the step of inputting the claim probability and the claim data in the fusion data into a pre-constructed anti-fraud scheme matching model to obtain the anti-fraud scheme corresponding to the target user comprises the steps of:
inputting the claim settlement probability into the anti-fraud scheme matching model to obtain an anti-fraud classification scheme;
matching the claim data with the claim data in the anti-fraud classification scheme to obtain a matching value;
and selecting the anti-fraud classification scheme with the maximum matching value as the anti-fraud scheme corresponding to the target user.
7. The method for analyzing fraud protection based on data fusion matching as claimed in claim 1, wherein the calculating feasibility of the fraud protection scheme according to the claim settlement probability, the claim settlement data and the claim settlement amount comprises:
determining the claim weight of the claim data by using a preset entropy algorithm;
determining a first feasibility value of the claim amount according to the claim weight;
determining a second feasibility value of the claim amount according to the claim settlement probability;
determining the feasibility of the anti-fraud scheme based on an average of the first feasibility value and the second feasibility value.
8. An anti-fraud analysis apparatus based on data fusion matching, the apparatus comprising:
the data fusion module is used for acquiring basic information data, credit data and claim settlement data of a historical user, and performing data fusion on the basic information data, the credit data and the claim settlement data to obtain historical fusion data;
the risk score value determining module is used for constructing a risk decision tree according to the historical fusion data and inputting the fusion data corresponding to the target user into the risk decision tree to obtain the risk score value of the target user;
the claim settlement probability calculation module is used for determining a fusion weight of the fusion data by using a preset analytic hierarchy process, and calculating the claim settlement probability of the target user according to the risk score value and the fusion weight through a preset dissimilarity weight algorithm;
the anti-fraud scheme determining module is used for inputting the claim settlement probability and the claim settlement data in the fused data into a pre-constructed anti-fraud scheme matching model to obtain an anti-fraud scheme corresponding to the target user;
and the anti-fraud scheme feasibility calculating module is used for determining the claim amount of the target user according to the claim compensation mode in the anti-fraud scheme and calculating the feasibility of the anti-fraud scheme according to the claim probability, the claim data and the claim amount.
9. An electronic device, characterized in that the electronic device comprises:
at least one processor; and (c) a second step of,
a memory communicatively coupled to the at least one processor; wherein, the first and the second end of the pipe are connected with each other,
the memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the method of anti-fraud analysis based on data fusion matching according to any one of claims 1 to 7.
10. A computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the method for anti-fraud analysis based on data fusion matching according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211591669.3A CN115809930A (en) | 2022-12-12 | 2022-12-12 | Anti-fraud analysis method, device, equipment and medium based on data fusion matching |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211591669.3A CN115809930A (en) | 2022-12-12 | 2022-12-12 | Anti-fraud analysis method, device, equipment and medium based on data fusion matching |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115809930A true CN115809930A (en) | 2023-03-17 |
Family
ID=85485645
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211591669.3A Pending CN115809930A (en) | 2022-12-12 | 2022-12-12 | Anti-fraud analysis method, device, equipment and medium based on data fusion matching |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115809930A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117291615A (en) * | 2023-11-27 | 2023-12-26 | 成都乐超人科技有限公司 | Visual contrast analysis method and device for overcoming anti-fraud based on network payment |
-
2022
- 2022-12-12 CN CN202211591669.3A patent/CN115809930A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117291615A (en) * | 2023-11-27 | 2023-12-26 | 成都乐超人科技有限公司 | Visual contrast analysis method and device for overcoming anti-fraud based on network payment |
| CN117291615B (en) * | 2023-11-27 | 2024-02-06 | 成都乐超人科技有限公司 | Visual contrast analysis method and device for overcoming anti-fraud based on network payment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110837931B (en) | Customer churn prediction method, device and storage medium | |
| CN107945024B (en) | Method for identifying internet financial loan enterprise operation abnormity, terminal equipment and storage medium | |
| CN110704572B (en) | Suspected illegal fundraising risk early warning method, device, equipment and storage medium | |
| CN111291816B (en) | Method and device for carrying out feature processing aiming at user classification model | |
| US20230004590A1 (en) | Ai-augmented auditing platform including techniques for automated adjudication of commercial substance, related parties, and collectability | |
| CN109711955B (en) | Poor evaluation early warning method and system based on current order and blacklist base establishment method | |
| CN110689438A (en) | Enterprise financial risk scoring method and device, computer equipment and storage medium | |
| CN111754116B (en) | Credit evaluation method and device based on label portrait technology | |
| CN107704512A (en) | Financial product based on social data recommends method, electronic installation and medium | |
| CN111260189B (en) | Risk control method, risk control device, computer system and readable storage medium | |
| CN108629685A (en) | loan product attribute determining method and server | |
| CN115545886A (en) | Overdue risk identification method, overdue risk identification device, overdue risk identification equipment and storage medium | |
| CN115809930A (en) | Anti-fraud analysis method, device, equipment and medium based on data fusion matching | |
| CN109146667B (en) | Method for constructing external interface comprehensive application model based on quantitative statistics | |
| CN113918718A (en) | Vehicle insurance user classification method, device, equipment and medium based on artificial intelligence | |
| CN113554504A (en) | Vehicle loan wind control model generation method and device and scoring card generation method | |
| CN112950347A (en) | Resource data processing optimization method and device, storage medium and terminal | |
| CN116843481A (en) | Knowledge graph analysis method, device, equipment and storage medium | |
| CN116468547A (en) | Credit card resource allocation method and system based on data mining | |
| CN113705201B (en) | Text-based event probability prediction evaluation algorithm, electronic device and storage medium | |
| CN113435746B (en) | User workload scoring method and device, electronic equipment and storage medium | |
| CN114708073A (en) | Intelligent detection method and device for surrounding mark and serial mark, electronic equipment and storage medium | |
| CN114626863A (en) | Detection method, device, equipment and storage medium for export tax cheating enterprise | |
| CN112419008A (en) | Automatic grading early warning method and device, electronic equipment and storage medium | |
| CN112926892A (en) | Capital matching method and device, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |