Contract risk analysis and early warning system and method and storage medium

Abstract

A contract risk analysis and early warning system, method and storage medium, comprising a data access module, which receives various contract related data in real time and pre-processes the contract related data; a feature extraction module, which extracts the multi-dimensional features of the contract related data; a risk prediction module, which determines the first association relationship between the contract related data corresponding to the anonymous feature and other contract related data, the second association relationship between the contract related data corresponding to the anonymous feature, and the third association relationship between the other contract related data, and performs risk prediction based on the association relationship and the multi-dimensional features. The method is implemented by using the above system, and the computer program of the storage medium is executed by the processor to realize the above method. The contract risk analysis and early warning system, method and storage medium can dynamically and batchly predict the contract damage risk, and classify and predict the contract, with high prediction accuracy, so that the enterprise can timely take risk control measures.

Description

Technical Field

[0001] This invention relates to the field of risk assessment technology, and in particular to a contract risk analysis and early warning system, method and storage medium. Background Technology

[0002] In business activities and legal practice, contracts are the core basis for establishing rights and obligations. To prevent potential losses during contract performance, contract risk assessment systems are widely used in corporate legal and risk control management. Currently, most contract risk assessment methods are based on static analysis models. Specifically, these systems typically rely on users manually inputting key contract elements (such as contracting parties, contract amount, performance period, and breach of contract clauses), and then comparing this information with a pre-set risk rule base or historical case database to generate a risk report at a fixed point in time.

[0003] However, those skilled in the art have found in practice that existing static assessment mechanisms have significant technical limitations. Due to the highly dynamic and time-sensitive nature of external variables such as the business environment, laws and regulations, market exchange rates, supply chain status, and the creditworthiness of the contracting party, risk assessment results generated based on information from a single point in time are easily diminished or rendered ineffective during contract performance. For example, a procurement contract initially assessed as low-risk may become high-risk due to subsequent changes in export control policies or a downgrade of the counterparty's credit rating, but existing systems cannot automatically capture these changes, leading to underreporting of risks. Furthermore, when conducting batch risk assessments of existing contracts, existing technologies often require manual re-entry and full data cleaning, resulting in low assessment efficiency and an inability to reflect real-time risk evolution trends.

[0004] Therefore, existing one-time contract risk assessment methods based on static input lack the ability to continuously track the time dimension and adapt to the dynamic external environment. They suffer from technical problems such as lagging assessment results, inability to reflect the dynamic evolution of risks, and inability to achieve real-time early warning. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to provide a contract risk analysis and early warning system, method and storage medium that can dynamically and in batches predict the risk of contract damage, classify contracts for prediction, and have high prediction accuracy, so that enterprises can take timely risk control measures.

[0006] This invention provides a contract risk analysis and early warning system, comprising:

[0007] The data access module is used to receive various contract-related data in real time and preprocess the contract-related data.

[0008] The feature extraction module is used to extract multidimensional features from the contract-related data, wherein the multidimensional features include at least named features and unnamed features;

[0009] The risk prediction module is used to determine the first correlation between the contract-related data corresponding to the unnamed feature and other contract-related data, the second correlation between the contract-related data corresponding to the unnamed feature and the third correlation between other contract-related data, and to perform risk prediction based on the first correlation, the second correlation, the third correlation and the multidimensional feature.

[0010] In one embodiment, the data access module further includes:

[0011] The data acquisition submodule is used to connect with the sales system, delivery system, contract management system, financial accounting system, and invoice system to obtain the contract-related data.

[0012] The preprocessing submodule is used to perform data deduplication, missing value filling, format standardization, entity alignment, and master data association processing on the obtained contract-related data.

[0013] In one embodiment, the feature extraction module includes:

[0014] The first feature extraction submodule is used to determine named and unnamed features based on the contract-related data.

[0015] The second feature extraction submodule is used to obtain customer features, contract amount features, payment period features, delivery node features, payment frequency features, historical performance features, industry features, and overdue features based on the contract-related data.

[0016] In one embodiment, the first feature extraction submodule further includes:

[0017] The initial screening unit is used to perform initial screening of contract-related data based on the contract name.

[0018] The comparison unit is used to extract the first core obligation data from the contract-related data, compare the first core obligation data with the named contract features, and obtain whether the feature corresponding to the contract-related data is a named feature or an unnamed feature.

[0019] In one embodiment, the risk prediction module further includes:

[0020] The first association submodule is used to determine the first association relationship between the contract-related data corresponding to the unnamed feature and other contract-related data, and the second association relationship between the contract-related data corresponding to the unnamed feature, based on the text information, business logic and performance behavior of the contract-related data.

[0021] The second association submodule is used to determine the third association between the other contract-related data based on the text information, data fields, and business logic of the other contract-related data.

[0022] The risk assessment submodule is used to predict risks based on the first correlation, the second correlation, the third correlation, and multidimensional features.

[0023] In one embodiment, the first association submodule further includes:

[0024] The extraction unit is used to extract the obligation dataset of the contract-related data corresponding to the unnamed feature based on the text information;

[0025] The core comparison unit is used to filter out the second core obligation data in the obligation dataset based on the corresponding business logic and obtain the feature label corresponding to the second core obligation data.

[0026] The dependency unit is used to obtain the dependency relationship between the contract-related data corresponding to the unnamed feature and other contract-related data, as well as the dependency relationship between the contract-related data corresponding to the unnamed feature, based on the text information, business logic, and performance behavior.

[0027] In one embodiment, the risk assessment submodule further includes:

[0028] The first evaluation unit is used to provide static evaluation results based on multidimensional features, feature labels, and second core obligation data;

[0029] The second evaluation unit is used to provide dynamic evaluation results based on the first association, the second association, the third association, and the multidimensional features.

[0030] In one embodiment, the system further includes:

[0031] The report generation module is used to generate reports related to contract damage rates based on risk prediction results.

[0032] This invention also provides a method for contract risk analysis and early warning, implemented using the system described above, and comprising the following steps:

[0033] Receive various contract-related data and preprocess the contract-related data;

[0034] Multidimensional features are extracted from the contract-related data, wherein the multidimensional features include at least named features and unnamed features;

[0035] Determine the first association between the contract-related data corresponding to the unnamed feature and other contract-related data, the second association between the contract-related data corresponding to the unnamed feature and other contract-related data, and the third association between them. Based on the first association, the second association, the third association and the multidimensional feature, perform risk prediction.

[0036] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method described above.

[0037] The contract risk analysis and early warning system, method and storage medium provided by this invention can dynamically and in batches predict the risk of contract damage, and classify contracts for prediction with high accuracy, so that enterprises can take timely risk control measures. Attached Figure Description

[0038] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0039] Figure 1 This is a system block diagram of the contract risk analysis and early warning system provided by the present invention.

[0040] Figure 2 This is a flowchart illustrating the contract risk analysis and early warning method provided by the present invention. Detailed Implementation

[0041] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some, not all, of the embodiments of the present invention. Based on the description of the present invention, all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of the present invention.

[0042] In the description of this invention, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can understand the specific meaning of the above terms according to the specific circumstances.

[0043] The terms “upper,” “lower,” “left,” “right,” “front,” “back,” “top,” “bottom,” “inner,” and “outer,” etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product of the invention is in use. They are only for the convenience of description and simplification, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the present invention.

[0044] The terms “first,” “second,” “third,” etc., are used merely to distinguish elements with similar properties, not to indicate or imply relative importance or a specific order.

[0045] The terms “include,” “comprising,” or any other variation thereof are intended to cover non-exclusive inclusion, which includes not only the elements listed but also other elements not expressly listed.

[0046] Example 1

[0047] Please see Figure 1 The contract risk analysis and early warning system provided by this invention includes:

[0048] The data access module is used to receive various contract-related data and preprocess the contract-related data.

[0049] It is understandable that the above modules further include:

[0050] The data acquisition submodule is used to connect with the sales system, delivery system, contract management system, financial accounting system, and invoice system to obtain contract-related data.

[0051] Understandably, the sales system contains data such as order amounts and customer credit data. Customer credit data includes external and internal credit data. External credit data mainly relies on the customer's basic information and business status, while internal credit data is a customer score given by the company based on the customer's creditworthiness. The delivery system includes data such as the company's current delivery progress, production progress, material status, customer return and exchange status, delivery cycle, and material price fluctuation status. The contract management system includes contract type (named or unnamed contract), performance status, text information, and change records. The financial accounts system includes payment records and aging information. The invoice system mainly includes payment information and information on both parties involved in the payment and receipt.

[0052] The preprocessing submodule is used to perform data deduplication, missing value filling, format standardization, entity alignment, and master data association processing on the obtained contract-related data.

[0053] Understandably, data deduplication involves deleting duplicate or redundant data records; missing value imputation involves filling in missing contract signing dates and customer information; format standardization involves standardizing text descriptions, contract numbers, and service names; entity alignment involves aligning the names of the same customer within the same system; and master data association mainly involves unifying the terminology used in different systems. After the above preprocessing, contract-related data can be well standardized.

[0054] The feature extraction module is used to extract multidimensional features from contract-related data. These multidimensional features include at least named features and unnamed features.

[0055] It can be understood that the above module may further include:

[0056] The first feature extraction submodule is used to determine named and unnamed features based on contract-related data.

[0057] The first feature extraction submodule mentioned above may include:

[0058] The initial screening unit is used to perform initial screening of contract-related data based on the contract name.

[0059] The comparison unit is used to extract the first core obligation data from the contract-related data, compare the first core obligation data with the named contract features, and obtain whether the features corresponding to the contract-related data are named features or unnamed features.

[0060] Understandably, for the sales industry, contracts mainly focus on named contracts, with a small number of unnamed contracts. Therefore, the main classification of contract-related data is named features and unnamed features. The determination of named and unnamed features can be based on the fields and tags in the text information of the contract-related data. The determination of named and unnamed features can also be based on the contract name and the first core obligation data. Contracts whose names and core obligation data match the named contract features are identified as named contracts. Contracts whose names do not match but whose core obligation data matches are also identified as named contracts. Contracts whose names match but whose core obligation data does not, whose names and core obligation data do not match, or whose core obligation data matches two or more named contracts are all identified as unnamed contracts. The initial screening based on the contract name can quickly compare the first core obligation data with the corresponding named contracts.

[0061] The second feature extraction submodule is used to obtain customer features, contract amount features, payment period features, delivery node features, payment frequency features, historical performance features, industry features, and overdue features based on contract-related data.

[0062] Understandably, customer characteristics may include the customer's company operating status, credit status, and asset status; contract amount characteristics may include total contract amount, unit price, prepayment, progress payment, taxes, warranty payment, penalties, and material price fluctuations; payment term characteristics may include payment term start date, payment term duration, payment deadline, payment method, and consequences of overdue payments; delivery node characteristics may include data such as delivered goods, delivery time, delivery location, delivery standards, and delivery vouchers; payment frequency characteristics may include payment cycle, number of payments, distribution of payment amounts, and payment triggering conditions; historical performance characteristics may include parameters such as on-time delivery rate, on-time payment rate, number of delayed days, number of delayed transactions, return rate, and payment in full rate; industry characteristics may include policy dependence, transaction habits, volatility with economic cycles, and financial characteristics; and overdue characteristics may include the number of overdue transactions, overdue rate, and overdue interval.

[0063] The risk prediction module is used to determine the first correlation between contract-related data corresponding to unnamed features and other contract-related data, the second correlation between contract-related data corresponding to unnamed features and other contract-related data, and the third correlation between other contract-related data. Risk prediction is performed based on the first correlation, second correlation, third correlation and multi-dimensional features.

[0064] It is known that the risk prediction module may further include:

[0065] The first association submodule is used to determine the first association between contract-related data corresponding to unnamed features and other contract-related data, and the second association between contract-related data corresponding to unnamed features, based on the text information, business logic, and performance behavior of the contract-related data.

[0066] It is understandable that the first relationship submodule may further include:

[0067] The extraction unit is used to extract the obligation dataset of contract-related data corresponding to unnamed features based on text information.

[0068] It is known that, depending on the type of obligation, the obligation dataset includes pure unnamed contracts, mixed unnamed contracts, and quasi-mixed contracts. Pure unnamed contracts can be understood as contracts that are unrelated to any named contracts and are not stipulated by law at all. Mixed unnamed contracts can be understood as contracts that are composed of the contents of at least two named contracts. Quasi-mixed contracts can be understood as contracts that are mainly composed of the contents of named contracts and contain some special contents that do not belong to named contracts.

[0069] The core comparison unit is used to filter out the second core obligation data in the obligation dataset based on the corresponding business logic and obtain the feature label corresponding to the second core obligation data.

[0070] It is known that the obligation dataset is filtered based on business logic to obtain the second core obligation data. The filtering can be based on key core factors such as the irreplaceability of the business logic, the time-determining nature of the obligation, and the starting point of the substantial dependency chain. Irreplaceability can be understood as the performance of the obligation cannot be completely replaced by monetary compensation. Time-determining nature can be understood as the obligation must be performed before a specific time point, otherwise the purpose of the contract will fail. The starting point of the substantial dependency chain can be understood as the existence or performance of other obligations being based on this obligation. The second core obligation data is matched with named contracts to understand which type of named contract features the second core obligation data has a higher similarity to, and corresponding feature labels are assigned.

[0071] The dependency unit is used to obtain the dependency relationships between contract-related data corresponding to unnamed features and other contract-related data, as well as the dependency relationships between contract-related data corresponding to unnamed features, based on text information, business logic, and performance behavior.

[0072] It can be understood that other contract-related data can be understood as contract-related data corresponding to named features. The text information may contain textual data that clearly records the dependency relationship between unnamed contracts. In the absence of clear textual records, the business logic can be judged based on the starting point of the substantial dependency chain. The performance behavior can be judged based on the correlation between the performance time and the correlation between the performance order of the contract-related data. The dependency relationship and feature tags constitute the first and second relationship.

[0073] The second relationship submodule is used to determine the third relationship between other contract-related data based on the text information, data fields, and business logic of other contract-related data.

[0074] As we know, text information can be used to determine relationships directly based on explicitly recorded text, data fields can be used to determine structural relationships, and business logic can be judged in the same way.

[0075] The risk assessment submodule is used to predict risks based on the first, second, and third relationships and multidimensional features.

[0076] It is known that the risk assessment submodule may further include:

[0077] The first evaluation unit is used to provide static evaluation results based on multidimensional features, feature labels, and the second core obligation data.

[0078] Understandably, the first evaluation unit is implemented based on anomaly detection algorithms. For contract-related data corresponding to unnamed features, the corresponding named contract features are determined based on feature labels, and the corresponding first risk value of the named contract is determined. The first risk value is adjusted based on the second core obligation data to obtain the corresponding second risk value. The second risk value is then adjusted based on multidimensional features to obtain the corresponding third risk value.

[0079] The second evaluation unit is used to provide dynamic evaluation results based on the first correlation, the second correlation, the third correlation, and multidimensional features.

[0080] It is understandable that the second assessment unit can use a time series model for prediction and assessment. The time series model supports incremental learning and dynamic weight adaptive adjustment. It predicts the correlation risk between contract-related data based on the first, second and third correlation relationships, and outputs the prediction results based on the multi-dimensional features acquired in real time. The third risk value is adjusted based on the prediction results to obtain the final risk value.

[0081] In some embodiments, the system further includes:

[0082] The report generation module is used to generate reports related to contract damage rates based on risk prediction results.

[0083] Understandably, reports related to contract damage rates can include contract damage rate analysis reports, risk distribution charts, customer risk rankings, and overdue monitoring dashboards, and can support chart rendering and one-click export.

[0084] Example 2

[0085] Please see Figure 2 This embodiment provides a method for contract risk analysis and early warning, implemented using the aforementioned system, and includes the following steps:

[0086] Receive various contract-related data and preprocess the contract-related data;

[0087] Extract multidimensional features from contract-related data, where the multidimensional features include at least named features and unnamed features;

[0088] Determine the first association between contract-related data corresponding to unnamed features and other contract-related data, the second association between contract-related data corresponding to unnamed features and other contract-related data, and the third association between other contract-related data. Based on the first association, second association, third association and multidimensional features, perform risk prediction.

[0089] Example 3

[0090] This embodiment provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the above-described method.

[0091] As described above, the contract risk analysis and early warning system, method, and storage medium provided by this invention can dynamically and in batches predict the risk of contract damage, and classify contracts for prediction with high accuracy, so that enterprises can take timely risk control measures.

[0092] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the appended claims.

Claims

1. A contract risk analysis and early warning system, characterized in that, include: The data access module is used to receive various contract-related data in real time and preprocess the contract-related data. The feature extraction module is used to extract multidimensional features from the contract-related data, wherein the multidimensional features include at least named features and unnamed features; The risk prediction module is used to determine the first correlation between the contract-related data corresponding to the unnamed feature and other contract-related data, the second correlation between the contract-related data corresponding to the unnamed feature and the third correlation between other contract-related data, and to perform risk prediction based on the first correlation, the second correlation, the third correlation and the multidimensional feature.

2. The contract risk analysis and early warning system of claim 1, wherein, The data access module further includes: The data acquisition submodule is used to connect with the sales system, delivery system, contract management system, financial accounting system, and invoice system to obtain the contract-related data. The preprocessing submodule is used to perform data deduplication, missing value filling, format standardization, entity alignment, and master data association processing on the obtained contract-related data.

3. The contract risk analysis and warning system of claim 1, wherein, The feature extraction module includes: The first feature extraction submodule is used to determine named and unnamed features based on the contract-related data. The second feature extraction submodule is used to obtain customer features, contract amount features, payment period features, delivery node features, payment frequency features, historical performance features, industry features, and overdue features based on the contract-related data.

4. The contract risk analysis and warning system of claim 3, wherein, The first feature extraction submodule further includes: The initial screening unit is used to perform initial screening of contract-related data based on the contract name. The comparison unit is used to extract the first core obligation data from the contract-related data, compare the first core obligation data with the named contract features, and obtain whether the feature corresponding to the contract-related data is a named feature or an unnamed feature.

5. The contract risk analysis and warning system of claim 1, wherein, The risk prediction module further includes: The first association submodule is used to determine the first association relationship between the contract-related data corresponding to the unnamed feature and other contract-related data, and the second association relationship between the contract-related data corresponding to the unnamed feature, based on the text information, business logic and performance behavior of the contract-related data. The second association submodule is used to determine the third association between the other contract-related data based on the text information, data fields, and business logic of the other contract-related data. The risk assessment submodule is used to predict risks based on the first correlation, the second correlation, the third correlation, and multidimensional features.

6. The contract risk analysis and warning system of claim 5, wherein, The first association submodule further includes: The extraction unit is used to extract the obligation dataset of the contract-related data corresponding to the unnamed feature based on the text information; The core comparison unit is used to filter out the second core obligation data in the obligation dataset based on the corresponding business logic and obtain the feature label corresponding to the second core obligation data. The dependency unit is used to obtain the dependency relationship between the contract-related data corresponding to the unnamed feature and other contract-related data, as well as the dependency relationship between the contract-related data corresponding to the unnamed feature, based on the text information, business logic, and performance behavior.

7. The contract risk analysis and early warning system as described in claim 6, characterized in that, The risk assessment submodule further includes: The first evaluation unit is used to provide static evaluation results based on multidimensional features, feature labels, and second core obligation data; The second evaluation unit is used to provide dynamic evaluation results based on the first association, the second association, the third association, and the multidimensional features.

8. The contract risk analysis and early warning system as described in claim 1, characterized in that, The system also includes: The report generation module is used to generate reports related to contract damage rates based on risk prediction results.

9. A method for contract risk analysis and early warning, characterized in that, The system described in any one of claims 1 to 8 comprises the following steps: Receive various contract-related data and preprocess the contract-related data; Multidimensional features are extracted from the contract-related data, wherein the multidimensional features include at least named features and unnamed features; Determine the first association between the contract-related data corresponding to the unnamed feature and other contract-related data, the second association between the contract-related data corresponding to the unnamed feature and other contract-related data, and the third association between them. Based on the first association, the second association, the third association and the multidimensional feature, perform risk prediction.

10. A computer-readable storage medium, characterized in that, It stores a computer program that, when executed by a processor, implements the method as described in claim 9.

Images