Intelligent detection method and device for trade contract and medium

Abstract

The invention discloses an intelligent detection method and device for a trade contract and a medium, and relates to the technical field of computer information, and the method comprises the steps: collecting historical contract data, and carrying out the preprocessing of the historical contract data; according to a preset feature project, extracting association relationship features in the preprocessed historical contract data, and constructing a data set based on the association relationship features; the data set comprises a training set and a test set; according to the feature importance, carrying out weight distribution on the association relationship features in the training set, and training a pre-deployed support vector machine model through the training set; verifying the trained support vector machine model through the test set, and optimizing the support vector machine model according to a verification result; and performing false detection on the current trade contract through the optimized support vector machine model. Through a machine learning technology, a large amount of trade contract data can be rapidly processed, false trade behaviors are accurately identified, and the detection efficiency and accuracy are greatly improved.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to an intelligent detection method, device and medium for a trade contract. Background Art

[0002] Trade contracts are the core legal documents for all kinds of economic and trade activities. They clearly define the rights and obligations of both parties to the transaction, and cover key information such as the specifications, prices, delivery methods, and payment terms of goods or services. They are an important basis for ensuring the smooth conduct of transactions and safeguarding the legitimate rights and interests of both parties. With the increasing complexity of economic globalization and trade activities, the terms of trade contracts are becoming more and more rich and detailed, involving not only basic transaction terms, but also many complex terms such as intellectual property protection, liability for breach of contract, and dispute resolution mechanisms. Therefore, in the current trade industry, the identification and prevention of false trade is a severe challenge facing enterprises.

[0003] However, the traditional method of detecting trade contracts usually involves manual review by professionals or simple rule matching (such as OCR technology), which involves preliminary automated review of contract documents and comparison with standard templates to screen suspicious contracts. This method is greatly affected by the professional level, experience and subjective judgment of the reviewers. It is not only time-consuming and laborious, but also prone to omissions. It may ignore some potential risks and problems, causing legal risks and economic losses to enterprises. Summary of the invention

[0004] In order to solve the above problems, this application proposes an intelligent detection method for trade contracts, including:

[0005] Collecting historical contract data and preprocessing the historical contract data;

[0006] According to the preset feature engineering, the correlation features in the preprocessed historical contract data are extracted, and a data set is constructed based on the correlation features; the data set includes a training set and a test set;

[0007] According to feature importance, weights are assigned to the association relationship features in the training set, and a pre-deployed support vector machine model is trained using the training set;

[0008] Verifying the trained support vector machine model through the test set, and optimizing the support vector machine model according to the verification result;

[0009] The optimized support vector machine model is used to detect falsehoods in current trade contracts.

[0010] On the other hand, the present application also proposes an intelligent detection device for trade contracts, comprising:

[0011] at least one processor; and,

[0012] a memory communicatively connected to the at least one processor; wherein,

[0013] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute an intelligent detection method for a trade contract as described in the above example.

[0014] On the other hand, the present application also proposes a non-volatile computer storage medium storing computer executable instructions, wherein the computer executable instructions are configured as: an intelligent detection method for a trade contract as described in the above example.

[0015] The intelligent detection method of trade contracts proposed in this application can bring the following beneficial effects:

[0016] By extracting the correlation features from the multi-dimensional data of both parties to the contract and constructing a data set, we can provide a comprehensive and valuable basis for risk assessment. On this basis, we can train and optimize the model. The generated risk assessment model can more accurately predict trade contract risks and provide scientific support for corporate decision-making.

[0017] Through machine learning technology, a large amount of trade contract data can be quickly processed. Compared with traditional manual review or simple rule matching, it can accurately extract key information from complex data, accurately identify false trade behaviors, greatly improve detection efficiency and accuracy, and reduce the heavy workload and human errors of manual review. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0019] Figure 1 A flowchart of an intelligent detection method for a trade contract in an embodiment of the present application;

[0020] Figure 2 This is a schematic diagram of the memory isolation process in an embodiment of the present application;

[0021] Figure 3 This is a schematic diagram of screening for equal interest relationships in the embodiments of this application;

[0022] Figure 4 A schematic diagram of a data set constructed in an embodiment of the present application;

[0023] Figure 5 This is a schematic diagram of the verification results of the verification set in the embodiment of this application;

[0024] Figure 6 This is a discount diagram of the verification result of the verification set in the embodiment of this application;

[0025] Figure 7 This is a schematic diagram of the model deployment in the embodiment of this application;

[0026] Figure 8 This is a schematic diagram of the user interface in the embodiment of the present application;

[0027] Fig. 9 This is a schematic diagram of an intelligent detection device for a trade contract in an embodiment of the present application. DETAILED DESCRIPTION

[0028] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.

[0029] The technical solutions provided by various embodiments of the present application are described in detail below in conjunction with the accompanying drawings.

[0030] like Figure 1 As shown, the embodiment of the present application provides an intelligent detection method for a trade contract, including:

[0031] S101: Collect historical contract data and pre-process the historical contract data.

[0032] Specifically, historical contract data is collected from the contract management system, financial system and business operation records of the enterprise based on the contract subject, including contract number, contract signing date, contract amount, payment direction, enterprise subject information (such as enterprise name, unified social credit code, registered address, office address, legal representative, shareholder information, key management personnel, etc.), specific transaction data between enterprises (such as transaction time, type, quantity, price of transaction products or services, etc.). Among them, the contract subject includes the contract party and the contract counterparty, and the historical contract data includes the contract party data and the contract counterparty data.

[0033] Furthermore, according to the direction of payment and collection, the contractual parties are divided into upstream parties and downstream parties, and the historical contract data are preprocessed, including data cleaning, data conversion and other processing. Data cleaning includes missing value processing, outlier processing and duplicate value processing; data conversion includes data standardization, data normalization and text data processing.

[0034] It should be noted that in the process of trade transactions, the flow of funds and the flow of goods or services are usually opposite. The upstream party refers to the party at the supply end of the supply chain, that is, the contractual counterparty that provides goods or services. In the transaction process, the upstream enterprise provides products or services to other enterprises and collects payment from the downstream enterprise. The downstream party refers to the party at the demand end of the supply chain, that is, the contractual counterparty that receives goods or services. The downstream enterprise purchases products or services from the upstream enterprise and pays the corresponding fees.

[0035] S102: According to preset feature engineering, extract association relationship features from the preprocessed historical contract data, and construct a data set based on the association relationship features; the data set includes a training set and a test set.

[0036] Specifically, the existence of equal interest relationships between the contract parties is determined, and based on the equal interest relationships, the association relationship features in the pre-processed historical contract data are extracted. Among them, the equal interest relationship includes at least one or more of the following: the corporate affiliation of the contract counterparty is consistent, the control relationship of the contract counterparty's affiliated enterprises is related, the equity of the contract counterparty's affiliated enterprises is cross-connected, the personnel management of the contract counterparty's affiliated enterprises overlaps, and the address information of the contract counterparty's affiliated enterprises is similar; the association relationship features corresponding to the equal interest relationship include: corporate affiliation association features, control relationship association features, equity cross-association features, personnel management association features, and address information association features.

[0037] It should be noted that if Figure 2 As shown, the interest relationship of consistent affiliation of the contractual counterparties' enterprises means that the upstream and downstream are the same enterprise; the interest relationship of related control relationship of the contractual counterparties' enterprises means that the upstream and downstream are parent and subsidiary companies or are controlled by the same actual controller; the interest relationship of cross-connected equity of the contractual counterparties' enterprises means that the upstream and downstream enterprises have cross-holdings; the interest relationship of overlapping personnel management of the contractual counterparties' enterprises means that the main responsible persons, directors, supervisors, and senior management personnel of the upstream and downstream enterprises are the same; the interest relationship of convergent address information of the contractual counterparties' enterprises means that the registered address, actual office address, business contact person or contact number of the upstream and downstream enterprises are the same.

[0038] Furthermore, the values ​​corresponding to the equity cross-correlation feature and the personnel management correlation feature are set as feature values, and based on a preset coding format (such as one-hot coding or label coding), the enterprise ownership correlation feature, the control relationship correlation feature and the address information correlation feature are encoded, and the encoded values ​​are set as the corresponding feature values. Obtain non-feature information in historical contract data, such as Figure 3 As shown, including fields such as the contract party, upstream contract counterparty, upstream contract counterparty material number, downstream contract counterparty, and downstream contract counterparty material number, the characteristic values ​​corresponding to the association relationship characteristics and the non-characteristic information in the historical contract data are integrated to construct a data set.

[0039] Among them, corresponding false labels are added to the confirmed false contract data in the data set.

[0040] S103: weighting the association relationship features in the training set according to feature importance, and training the pre-deployed support vector machine model using the training set.

[0041] Specifically, the correlation coefficient between the association relationship feature and the trade contract is calculated by a statistical method. The closer the absolute value of the correlation coefficient is to 1, the stronger the correlation between the feature and the target variable is, and the higher the importance of the feature may be. The feature importance of the association relationship feature is determined according to the correlation coefficient, and the corresponding weight is assigned to the association relationship feature in the training set according to the feature importance, and multiple feature weight combinations are set based on the weight. In the embodiment of the present application, the variance analysis method can be used to analyze the relationship between the categorical variable and the continuous variable.

[0042] It should be noted that the importance analysis of association relationship features can also be achieved through domain expert knowledge or experimental adjustment.

[0043] Furthermore, based on the feature weight combination, the corresponding training set data is input into the pre-deployed support vector machine model in sequence to obtain the classification result output by the support vector machine model, and the prediction error of the support vector machine model is obtained based on the classification result. According to the prediction error, the model parameters of the support vector machine model are adjusted through the optimization algorithm, and the support vector machine model is iteratively trained based on the training set until the prediction error converges.

[0044] During the model training process, the weights are adjusted through optimization algorithms to minimize the model's prediction error. First, the model parameters are initialized and a loss function such as mean square error is selected to measure the difference between the model prediction and the actual value; secondly, the optimization algorithm (such as gradient descent) is applied to calculate the gradient to update the model parameters, and training is stopped when the performance is no longer improved; finally, the model parameters are continuously adjusted to minimize the prediction error and improve the prediction accuracy of the model on new data.

[0045] It should be noted that the identification of false trade may involve complex patterns and boundaries, and the clustered support vector machine can capture these complexities through nonlinear kernel techniques; the task is a binary classification problem (whether the contract is a false trade), and the support vector machine performs well in binary classification problems; the support vector machine finds the support vector on the decision boundary through optimization, which means that it naturally gives certain features greater weights; in the present invention, this can help the model identify which features are most critical for identifying false trade; by integrating multiple support vector machines, the clustered support vector machine can improve the generalization ability of the model and reduce the possible overfitting problem of a single model.

[0046] S104: verifying the trained support vector machine model using the test set, and optimizing the support vector machine model according to the verification result.

[0047] Specifically, the input feature data in the test set is input into the trained support vector machine model. The model will classify and predict each sample according to its learned decision boundary and output the prediction results. Based on the prediction results, the prediction performance of the support vector machine model is verified through evaluation indicators.

[0048] Among them, the evaluation indicators include accuracy, precision, recall, standard error, etc.

[0049] Furthermore, based on the validation results, the model optimization strategy is implemented and the model parameters are adjusted.

[0050] It should be noted that accuracy is the most intuitive performance indicator, which indicates the proportion of samples correctly predicted by the model to the total number of samples. In some cases, a high accuracy rate means better model performance. Precision, also known as the positive prediction value rate, measures the proportion of samples predicted by the model as positive that are actually positive. For unbalanced data sets, precision is more important than accuracy. Recall rate measures the proportion of samples correctly predicted as positive among all actual positive samples. Recall rate is particularly important in scenarios where it is necessary to identify as many positive samples as possible. For example, Figure 4 and Figure 5 As shown in Figure 1, RMSE is the square root of MSE, also known as standard error. It has the same unit as the original data, which makes it more intuitive when interpreting model errors. The smaller the RMSE, the more accurate the model's predictions.

[0051] S105: Perform false detection on the current trade contract through the optimized support vector machine model.

[0052] Specifically, the current association relationship features and current non-feature data in the current trade contract are extracted, the current association relationship features and current non-feature data are input into the optimized support vector machine model, the prediction results of the support vector machine are obtained, and according to the prediction results, it is determined whether the current trade contract is a false trade contract.

[0053] It should be noted that if Figure 6 As shown in Figure 2, after the support vector machine model optimization converges, it is deployed into the actual contract review system to automatically identify suspected fraudulent trade. The model is integrated into the existing contract review system to ensure that it can receive input data, make predictions, and return results; Figure 7As shown, design a user-friendly interface that allows customer personnel to easily import and use model outputs, and supplement and iterate features as needed. After deployment, continuously monitor the performance of the model in the actual environment to ensure that it can adapt to new data and possible changes. Establish a data pipeline to ensure that the model can receive the latest and accurate contract data. Implement a feedback mechanism that allows users to report any problems or misjudgments with the model so that the model can be continuously improved. Ensure that the deployment of the model complies with all relevant laws, regulations, and data protection standards. Consider the scalability of the model to ensure that the system can handle increased data volumes and user needs. Establish a maintenance plan to regularly update the model to adapt to new data and business needs. Provide detailed documentation and training to help reviewers understand how the model works and how to use it. As Figure 8 As shown, after deployment, the performance of the model is continuously monitored and updated and maintained based on new data and feedback.

[0054] By extracting the correlation features from the multi-dimensional data of both parties to the contract and constructing a data set, we can provide a comprehensive and valuable basis for risk assessment. On this basis, we can train and optimize the model. The generated risk assessment model can more accurately predict trade contract risks and provide scientific support for corporate decision-making.

[0055] Through machine learning technology, a large amount of trade contract data can be quickly processed. Compared with traditional manual review or simple rule matching, it can accurately extract key information from complex data, accurately identify false trade behaviors, greatly improve detection efficiency and accuracy, and reduce the heavy workload and human errors of manual review.

[0056] like Fig. 9 As shown, the embodiment of the present application also proposes an intelligent detection device for trade contracts, including:

[0057] at least one processor; and,

[0058] a memory communicatively connected to the at least one processor; wherein,

[0059] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute an intelligent detection method for a trade contract as described in any of the above embodiments.

[0060] An embodiment of the present application further provides a non-volatile computer storage medium storing computer executable instructions, wherein the computer executable instructions are configured as: an intelligent detection method for a trade contract as described in any of the above embodiments.

[0061] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device and medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiments.

[0062] The devices and media provided in the embodiments of the present application correspond one-to-one to the methods. Therefore, the devices and media also have similar beneficial technical effects as the corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the devices and media will not be repeated here.

[0063] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0064] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0065] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0066] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0067] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0068] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0069] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0070] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0071] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. An intelligent detection method for trade contracts, characterized in that: include: Collecting historical contract data and preprocessing the historical contract data; According to the preset feature engineering, the correlation features in the preprocessed historical contract data are extracted, and a data set is constructed based on the correlation features; The data set includes a training set and a test set; According to feature importance, weights are assigned to the association relationship features in the training set, and a pre-deployed support vector machine model is trained using the training set; Verifying the trained support vector machine model through the test set, and optimizing the support vector machine model according to the verification result; The optimized support vector machine model is used to detect falsehoods in current trade contracts.

2. The intelligent detection method of a trade contract according to claim 1 is characterized in that: The collecting of historical contract data and preprocessing of the historical contract data specifically includes: Based on the contract subject, historical contract data is collected, wherein the historical contract data includes the contract party data and the contract counterparty data; the contract subject includes the contract counterparty and the contract party; According to the direction of payment and receipt, the contractual counterparties are divided into upstream parties and downstream parties.

3. The intelligent detection method of a trade contract according to claim 2 is characterized in that: The extracting of association relationship features from the pre-processed historical contract data according to the preset feature engineering specifically includes: Determine the equal interest relationship between the contract parties, and based on the equal interest relationship, extract the association relationship features in the pre-processed historical contract data; The equal interest relationship includes at least one or more of the following: The enterprises of the contract counterparty are consistent, the control relationship of the enterprises of the contract counterparty is related, the equity of the enterprises of the contract counterparty is cross-connected, the personnel management of the enterprises of the contract counterparty is overlapping, and the address information of the enterprises of the contract counterparty is similar; The association relationship characteristics corresponding to the equal interest relationship include: enterprise affiliation association characteristics, control relationship association characteristics, cross-equity association characteristics, personnel management association characteristics, and address information association characteristics.

4. The intelligent detection method of a trade contract according to claim 3 is characterized in that: The step of constructing a data set based on the association relationship features specifically includes: Setting the numerical values ​​corresponding to the equity cross-correlation feature and the personnel management correlation feature as feature values; Based on a preset coding format, the enterprise affiliation association feature, the control relationship association feature and the address information association feature are encoded, and the encoded values ​​are set as corresponding feature values; The feature value corresponding to the association relationship feature and the non-feature information in the historical contract data are integrated to construct a data set.

5. The intelligent detection method of a trade contract according to claim 1 is characterized in that: The step of allocating weights to the association relationship features in the training set according to the feature importance specifically includes: Calculate the correlation coefficient between the association relationship feature and the trade contract by statistical methods, and determine the feature importance of the association relationship feature according to the correlation coefficient; According to the feature importance, corresponding weights are assigned to the association relationship features in the training set, and multiple feature weight combinations are set based on the weights.

6. The intelligent detection method of a trade contract according to claim 5, characterized in that: The training of the pre-deployed support vector machine model by using the training set specifically includes: Based on the feature weight combination, the corresponding training set data is sequentially input into the pre-deployed support vector machine model; Obtaining a classification result output by the support vector machine model, and obtaining a prediction error of the support vector machine model based on the classification result; According to the prediction error, the model parameters of the support vector machine model are adjusted by an optimization algorithm, and the support vector machine model is iteratively trained based on the training set until the prediction error converges.

7. The intelligent detection method of a trade contract according to claim 1, characterized in that: The optimized support vector machine model is used to detect the current trade contract for false information, specifically including: Extracting current association relationship features and current non-feature data in the current trade contract, and inputting the current association relationship features and the current non-feature data into the optimized support vector machine model; The prediction result of the support vector machine is obtained, and according to the prediction result, it is determined whether the current trade contract is a false trade contract.

8. The intelligent detection method of a trade contract according to claim 1 is characterized in that: After constructing the data set based on the association relationship features, the method further includes: Acquire the false contract data in the data set, and add a false label to the false contract data.

9. An intelligent detection device for trade contracts, characterized in that: include: at least one processor; as well as, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute an intelligent detection method for a trade contract as described in any one of claims 1 to 8.

10. A non-volatile computer storage medium storing computer executable instructions, characterized in that: The computer executable instructions are configured as: an intelligent detection method for a trade contract as described in any one of claims 1 to 8.

Images