Dynamic credit scoring method and system based on agricultural whole industry chain behavior data

Abstract

The invention provides a dynamic credit scoring method and system based on agricultural whole industry chain behavior data. The method comprises the steps of collecting agricultural technology learning and application behaviors, industry chain transaction and operation data and environment and time sequence data of a user; the scoring index weight is dynamically adjusted through a seasonal weight adaptive module according to an agricultural hour calendar, the user index and the average level of the same category of farmers in the same region are relatively evaluated through a regional correction module, and the agricultural material and agricultural machine data and the operation output are associated through a deep causal analysis module to calculate the use efficiency; and finally, outputting a credit score by using the credit scoring model. The system comprises corresponding function modules. According to the method, the agricultural industry rule is deeply fused into the credit model for the first time, the technical problems that a traditional method cannot quantify the operation capability, is not suitable for agricultural seasonal and regionality and cannot establish input-output causal association are solved, and more accurate and more economical credit evaluation of an agricultural operation subject is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of credit scoring, in particular to a dynamic credit scoring method based on agricultural full-industry-chain behavior data. BACKGROUND

[0002] Credit scoring is the core technology of financial technology, which predicts default risk by constructing statistical models through legal collection of user's identity, assets, historical credit records and other data. With the development of big data, some solutions begin to introduce e-commerce transactions, social behavior and other data.

[0003] In related technologies, such as e-commerce platform transaction flow data scoring solutions, loans are provided by analyzing the sales, order stability, customer evaluation and other data of merchants on e-commerce platforms. The steps of the solution include: 1. Collecting the historical transaction flow data of the user on the platform; 2. Calculate key indicators: monthly average sales, growth rate, refund rate, and praise rate; 3. Input the indicators into the preset model to output the credit score.

[0004] The above solution is based on the logic of urban commercial activities and is completely unsuitable for the agricultural production field which is dominated by natural laws and has long cycle and high risk characteristics. SUMMARY

[0005] The technical problem to be solved by the present application is how to reasonably score credit in the agricultural production field. The present application provides a dynamic credit scoring method based on agricultural full-industry-chain behavior data.

[0006] The dynamic credit scoring method based on agricultural full-industry-chain behavior data according to the embodiments of the present application comprises: S10, collecting multi-dimensional behavior data of the user, the multi-dimensional behavior data at least including: agricultural technology learning and application behavior data, industry chain transaction and operation data, environmental and time sequence data; S20, processing the collected multi-dimensional behavior data, comprising: Seasonal weight self-adaptation: dynamically adjusting the weights of different evaluation dimensions in the credit scoring model according to the preset agricultural calendar corresponding to the user; Regional correction: comparing the operation indicators of the user with the average level of farmers of the same region and the same product category in the database to obtain a relative evaluation score; Deep causal analysis: correlating the agricultural input data, agricultural machinery use data and the final agricultural product output data of the user to calculate the input efficiency and asset income generation capacity; S30, inputting the indicator data processed into the credit scoring model, calculating and outputting the credit score of the user.

[0007] According to some embodiments of the present application, the agricultural technology learning and application behavior data in step S10 includes: user's viewing and response records of agricultural risk early warning, learning and test completion records of agricultural technology courses, and consumption records of agricultural materials or agricultural machinery triggered after learning; In step S20, the viewing and response records, learning and test completion records, and consumption records are quantitatively evaluated as a continuous behavior chain.

[0008] In some embodiments of the present application, the seasonal weight adaptation includes: Identifying the current farming stage in which the user is located; Calling the weight configuration predefined for the farming stage, wherein in the fund investment period, the weight of the cash flow related indicators is reduced, and the weights of the production material preparation situation and the technology application related indicators are increased.

[0009] According to some embodiments of the present application, the industry chain transaction and operation data includes: first industry production data, second industry processing and order data, and third industry market and sales data; The method further includes: establishing a business association relationship between the first industry production data, the second industry processing and order data, and the third industry market and sales data, and evaluating the operation stability and the industry chain resilience of the user based on the association relationship.

[0010] In some embodiments of the present application, establishing the business association relationship includes: Identifying the matching degree between the first industry production standard and the quality requirement of the second industry procurement order; Identifying the adaptability between the second industry product output and the third industry sales channel and price.

[0011] According to some embodiments of the present application, the second industry processing and order data are collected by automatically identifying and structuring the uploaded procurement contract files through optical character recognition technology.

[0012] According to the dynamic credit scoring system based on agricultural full industry chain behavior data according to the embodiments of the present application, the dynamic credit scoring system is used to realize the dynamic credit scoring based on the agricultural full industry chain behavior data as described above, and the dynamic credit scoring system includes: A data acquisition module configured to acquire multi-dimensional behavior data of a user; A data processing module configured to process the multi-dimensional data, including performing seasonal weight adaptation, regional correction, and deep causal analysis; A model engine module configured to store and run a credit scoring model, taking the processed index data as input, and calculating a credit score; The application output module is configured to output the credit score, the level and the score report.

[0013] According to some embodiments of the present application, the dynamic credit scoring system further comprises: The scenario service packaging module is configured to match and provide scenario financial services corresponding to specific agricultural industry chain links for the user according to the credit scoring result of the user.

[0014] The electronic device according to the embodiments of the present application comprises a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the dynamic credit scoring method based on agricultural full industry chain behavior data as described above when executing the program.

[0015] The computer readable storage medium according to the embodiments of the present application has a computer program stored thereon, and the computer program is executed by a processor to implement the dynamic credit scoring method based on agricultural full industry chain behavior data as described above.

[0016] The present application has the following beneficial effects: The present application combines the dynamic behavior data of farmers' agricultural technology learning and industry chain operation with the characteristics of agricultural seasons and regions, establishes the causal relationship between agricultural input and operation output, and constructs a more accurate and inclusive credit scoring system, effectively solving the problems of high misjudgment rate and poor adaptability of traditional models in the agricultural field. It not only establishes a digital credit identity for farmers who lack credit records, but also scientifically guides the development of agricultural operation towards modernization and efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 FIG. 1 is a schematic diagram of a dynamic credit scoring system based on agricultural full industry chain behavior data according to an embodiment of the present application; Figure 2 FIG. 2 is a flowchart of a dynamic credit scoring method based on agricultural full industry chain behavior data according to an embodiment of the present application. DETAILED DESCRIPTION

[0018] In order to further illustrate the technical means and effects taken by the present application to achieve the predetermined purposes, the present application will be described in detail below in combination with the drawings and preferred embodiments.

[0019] The description of the method flow in the specification of the present application and the steps of the flowchart in the drawings of the present application do not necessarily strictly follow the step numbers, and the method steps can change the execution order. Moreover, some steps can be omitted, multiple steps can be combined into one step for execution, and / or one step can be divided into multiple steps for execution.

[0020] Traditional credit scoring methods have serious technical blind spots in the agricultural and pastoral scenes, and cannot accurately assess the true credit status of new agricultural business entities. The credit evaluation method proposed in the present invention solves the following core problems: 1. Unable to assess agricultural management capability: Existing models rely on financial historical data and cannot identify and quantify the core management capabilities of farmers such as "whether they can farm and manage", resulting in a large number of farmers with technology but lack of credit records being unable to obtain loans.

[0021] 2. Unable to adapt to the laws of the agricultural industry: general models are static and cannot understand the seasonality of agricultural production (such as cash flow tightness during the spring planting investment period), regionalism (different crops in different regions are not comparable), and accidental events (such as short-term cash flow interruption caused by marriage, funeral, and marriage).

[0022] 3. Unable to establish a deep causal relationship: the use of data such as agricultural materials and agricultural machinery is limited to the "asset list" level, and cannot analyze the causal relationship between "use efficiency" and "operational output".

[0023] 4. Unable to form a credit view covering the entire agricultural industry chain: traditional models have isolated data sources and cannot integrate the chain of business data from the primary industry (production), the secondary industry (processing), and the tertiary industry (circulation services), resulting in credit assessment like a blind man feeling an elephant, unable to assess the true value and risk of the business entity from the perspective of the entire industry ecosystem.

[0024] The system architecture of the present invention is shown in the accompanying Figure 1 The dynamic scoring process is shown in the accompanying Figure 2 The following describes the scheme of the present invention in two specific examples. It should be understood that the following description is only exemplary and should not be construed as a specific limitation of the present invention.

[0025] Example 1: This example takes a rice grower "Zhang" as an example.

[0026] User background: Zhang, who grows rice in a county in the south, has 50 mu of rice, is a registered user of the platform.

[0027] I. Data collection and index system; The system establishes a special data file for Zhang and collects the following multi-dimensional data and calculates the index: 1. Agricultural technology learning and application behavior ("Rain Agricultural Technology" dimension): Data: Zhang received a "early spring cold" warning from the platform in early March; he immediately learned the "water rice seedling period cold prevention technology" course and completed the test; after learning, he placed an order on the platform to purchase the recommended "cold prevention film".

[0028] Index calculation: Risk response rate = 1 (promptly check the warning); Technical course completion = 1 (complete learning and pass the test); Integration of learning and use = 1 (purchased related products after learning); This series of behaviors constitutes a complete "risk response" positive cycle, indicating that Zhang has good risk awareness and execution ability, and the related indicators have high scores.

[0029] 2. Industry chain transaction and operation data: Data: Zhang purchased special rice fertilizer of a specific brand this season; his rice was sold on the platform as "green and high-quality rice" after harvest, with a 15% higher price than ordinary rice; he also provided his harvesting machine for service on the platform.

[0030] Index calculation: Efficiency of agricultural input = total sales of rice / total purchase of agricultural materials = 120000 yuan / 30000 yuan = 4.0; High-value agricultural product sales ratio = sales of high-quality rice / total sales = 100%; Agricultural machinery service revenue generation ability = agricultural machinery service income / agricultural machinery asset value = 10000 yuan / 50000 yuan = 0.2; High input-output ratio and high-end product sales indicate efficient operation; agricultural machinery revenue generation indicates high asset utilization rate, which are strong credit characteristics.

[0031] 3. Environmental and time series data: Data: The platform determines Zhang's farming calendar based on his location and crops: sowing in March, nursery management in April, and harvesting in August.

[0032] Index calculation: Seasonal adaptation coefficient: In March (sowing period), the system checks Zhang's agricultural material preparation, and because he has prepared seeds, fertilizer, and cold protection film, this item scores high.

[0033] Regional production level index: Zhang's yield per mu is 600 kg, and the average yield per mu of local farmers of the same category is 550 kg, so his index is 600 / 550 ≈ 1.09 (higher than the average level).

[0034] The above multi-dimensional behavior data can be legally collected through authorization of the agricultural whole industry chain platform.

[0035] II. Detailed explanation of dynamic scoring process: Step S1: input preprocessed user data; The system pulls all the above raw data about Zhang, cleanses and integrates them.

[0036] Step S2: Seasonal weight self-adaptation; Assume the current time is March 15th. The model engine determines that Zhang is in the sowing period according to the "farming calendar".

[0037] Dynamic weight adjustment: the system automatically calls the "sowing period" weight configuration: Cash flow weight from 0.3 to 0.1 (understand that this is the input period, cash is normal).

[0038] Agricultural material preparation rate weight from 0.1 to 0.4 (focus on the preparation of production materials).

[0039] Technology application weight from 0.2 to 0.3 (focus on whether the sowing technology is in place).

[0040] This mechanism ensures that the scoring model does not mistakenly lower Zhang's credit score during the most cash-strapped sowing period, but instead focuses on whether he has made substantial preparations for spring farming.

[0041] Step S3: Regional correction; The system queries the average data of "rice growers in Zhang's county" in the database.

[0042] Relative evaluation: Zhang's per mu input cost is 1000 yuan, and the local average is 900 yuan. Simply looking at the absolute value, Zhang's cost is slightly higher. But combined with its per mu yield (600 kg) far higher than the average level (550 kg), and the product quality is better, the system will calculate that its unit cost output rate is significantly higher than the average level.

[0043] Avoiding "one-size-fits-all", Zhang's higher efficiency and output, and more refined operation, therefore, in the "operating ability" dimension, a higher relative score is obtained.

[0044] Step S4: Deep causal analysis; Agricultural material data causal analysis: The system executes the analysis of agricultural material input efficiency (Zhang ID, rice) function. This function will record the "special rice fertilizer" purchased by Zhang, and its final "green high-quality rice" sales and premium.

[0045] Conclusion: the system confirms that there is a strong causal relationship between the specific agricultural materials purchased by Zhang and the high-quality and high-priced output achieved. This proves that his input is "smart and efficient", rather than blind consumption.

[0046] Agricultural machinery data causal analysis: The system executes the evaluation of agricultural machinery usage value (Zhang ID) function. This function finds that its harvester is not only self-used, but also undertakes 20 external job services on the platform, generating 10000 yuan in revenue, with a usage rate of 80%.

[0047] Zhang's agricultural machinery is a "living asset that can generate money", not a "idle dead asset". Its ability to create cash flow is quantitatively evaluated.

[0048] This step converts simple "consumption records" and "asset lists" into deep insights into the quality of users' business decisions and the ability to operate assets, which traditional models cannot achieve.

[0049] Step S5: Comprehensive calculation and output; All index scores after seasonal adjustment, regional correction and causal analysis are input into the trained Gradient Boosting Decision Tree (GBDT) model.

[0050] Final output: After model comprehensive calculation, Zhang's credit score is 850 points, and the credit level is AA.

[0051] Score report: The report clearly lists key bonus items, such as: "Bonus for adopting disaster prevention technology and effective implementation", "Bonus for users with input-output efficiency 90% higher than regional average", "Bonus for agricultural machinery assets creating additional cash flow".

[0052] It is worth emphasizing that, due to the fact that the prior art scheme does not collect data such as farmers' production technology learning behavior, farming operation records, and environmental risk response, and does not consider the seasonal periodicity of agricultural production, the model cannot reflect the "operational stability" and "risk response ability" of farmers, resulting in a high misjudgment rate of agricultural user credit assessment.

[0053] Because the prior art scheme uses agricultural and machinery data only for purchase amount or asset ownership registration, it fails to deeply analyze the causal chain between them and the final agricultural product yield, quality, and sales price, resulting in the model being unable to distinguish between "blind investment" and "efficient investment", ultimately causing credit funds to flow to users with low operating efficiency.

[0054] Embodiment 2: This embodiment takes "Mr. Wang, head of the cooperative" as an example to show how three-industry integrated data drives credit assessment.

[0055] User background: Mr. Wang operates a vegetable cooperative engaged in planting (primary industry), primary processing (secondary industry), and e-commerce sales (tertiary industry).

[0056] I. Data collection and index system; The system establishes a special data file for Zhang and collects the following multi-dimensional data and calculates the indicators: First industry (production base) data: Data: 100 mu of green certified vegetable planting area, standardization production log, chemical fertilizer and pesticide use record, and soil moisture data collected by Internet of Things of Mr. Wang's cooperative.

[0057] Indicator calculation: production specification, environmental adaptability.

[0058] Second industry (value conversion) data: Data: Wang signed a long-term procurement order with a well-known clean vegetable processing plant; the order's quality control standards, price terms, and historical performance records; the cooperative's own primary processing equipment and capacity data.

[0059] Indicator calculation: order stability, processing value-added capacity.

[0060] Technical implementation: The platform automatically identifies and structures the uploaded procurement contracts through OCR (Optical Character Recognition) technology, ensuring data authenticity and traceability.

[0061] Third industry (value realization) data: Data: Wang's sales volume, repeat purchase rate, and customer review data on the e-commerce channels connected by the platform; the logistics delivery speed of his products.

[0062] Indicator calculation: market recognition, and repayment speed.

[0063] Technical implementation: Through API interface, the platform is connected with e-commerce platforms to obtain verifiable sales data in real time.

[0064] II. Detailed explanation of dynamic scoring process; Step S1: Input pre-processed user data; The system pulls all the above raw data about Wang, cleans, integrates, and correlates them. Key actions: The system automatically correlates Wang's "green vegetable" production records, "processing plant order" requirements, and "e-commerce sales" data through a data model to form a traceable data chain.

[0065] Step S2: Seasonal weight self-adaptation; Example: The current time is the planting season. The model engine determines that Wang is in the pre-production investment stage.

[0066] Dynamic weight adjustment: The system automatically calls the "pre-production investment stage" weight configuration, reducing the assessment weight of cash flow indicators and significantly increasing the assessment weight of production material preparation and downstream order certainty (from second industry data).

[0067] This mechanism ensures that the model will not deduct points due to Wang's financial difficulties during the planting season, but will give him points for holding stable processing plant orders (strong credit signal).

[0068] Step S3: Regional correction; Example: The system queries the average data of "similar vegetable cooperatives in Wang's region."

[0069] Relative evaluation: Wang's cost per mu may be slightly higher than the average, but his output value per unit area (which integrates the order price of secondary industry and the retail price of tertiary industry) is far higher than the average level.

[0070] The model accurately identified that Wang's "high investment" brought "high returns," and his operating efficiency was among the best in the local area.

[0071] In summary, this invention achieves three major synergistic effects within the existing architecture through innovation in data processing logic: Effect 1: The credit multiplication effect of the three-industry data chain; Technical implementation: In the data collection and indicator calculation layer, technologies such as OCR and API are used to link data from the primary, secondary and tertiary industries into a chain; in the credit model engine, algorithms have been specially developed to identify chain relationships such as "contract farming" and "integrated production, processing and sales".

[0072] This shifts credit assessment from a "single-point assessment" of the entity to an assessment of the "resilience of the industrial chain." A farmer with stable orders (secondary industry) faces far less risk than a farmer who blindly engages in production. This creates a risk assessment paradigm based on industrial ecosystem relationships.

[0073] Effect 2: The two-way empowerment of scenario-based services and credit scoring; Technical implementation: At the application output layer, the credit score results are directly used to unlock specific "scenario-based financial services", such as recommending "order financing" or "industrial chain upgrade loan" to Mr. Wang.

[0074] This breaks down the boundaries between "credit assessment" and "business services," establishing a positive feedback loop where "high-score credit unlocks high-quality scenario services, and the use of scenario services generates positive data to improve credit scores," thus building a virtuous cycle.

[0075] Effect 3: Risk penetration management brought about by end-to-end online processing; Technical Implementation: By utilizing technologies such as remote audio and video recording and electronic signatures, the entire process of pre-loan investigation, mid-loan review, and post-loan management was made online at the application layer without changing the system architecture.

[0076] This not only improves efficiency, but more importantly, it transforms previously offline and opaque business practices (such as face-to-face meetings and on-site inspections) into online, recordable, and analyzable structured data, enabling penetrating and digital management of risks, which is a significant advancement in risk control methods.

[0077] Because this invention is the first to systematically integrate dynamic behavioral data and industry characteristic logic from the entire agricultural industry chain into a credit model, and as shown in the embodiments, it can specifically identify users like "Zhang" and their good business habits and capabilities, thus having the following advantages compared to existing technologies: 1. More accurate assessment results: It no longer assesses "how much money a user has," but rather "how well a user knows how to make money." For example, it can accurately identify high-quality farmers like Mr. Zhang who are skilled in management and have development potential, thus reducing credit risk.

[0078] 2. More inclusive services: It establishes a "digital credit ID card" for a large number of new agricultural business entities that lack traditional collateral and credit history. For example, Mr. Zhang was able to obtain a loan based on his excellent "Rain Farm Technology" practices and efficient operational data, solving the "first mile" problem of agricultural financing.

[0079] 3. Positive Guiding Role for the Industry: Encourages users to actively learn agricultural techniques, invest scientifically, and operate with integrity. Just as Zhang's behavior directly improved his credit score, this will guide the entire industry towards modernization and efficiency.

[0080] Because this invention is not a simple superposition of existing technologies, but rather an organic system with synergistic effects and feedback loops, it achieves for the first time a deep coupling between credit assessment models and agricultural industry laws, thus resulting in the following prominent substantive features and significant technological advancements: 1. The risk management model has been upgraded: from "single-point" management of individual credit entities to comprehensive risk management of the "whole chain" of the "planting-processing-sales" industry, thus systematically reducing credit risk.

[0081] 2. It activates the value of agricultural data elements: By leveraging the data chain of the integration of primary, secondary, and tertiary industries, the previously dormant agricultural production, processing, and sales data are transformed into quantifiable and priced "data assets," ultimately enabling inclusive financial groups to "use data as an asset to enhance credit for loans."

[0082] 3. A closed-loop operation mechanism for the rural digital economy has been established: The platform not only provides credit, but also revitalizes rural assets and bridges the "last mile" of financial services by connecting all links of the industrial chain.

[0083] Finally, it should be emphasized that in the specific implementation of this invention, the collection, processing, and use of all user data strictly adhere to the principles of legality, legitimacy, and necessity, and user privacy has been fully protected through technical means and management systems. Specific measures are as follows: Authorization and Consent Mechanism: Before using this agricultural industry chain platform, users must register and explicitly agree to the platform's Terms of Service and Privacy Policy. This agreement clearly and explicitly informs users that their data will be used for credit assessment and has obtained the user's separate consent. All data collection is conducted within the scope authorized by the user.

[0084] Data anonymization and masking: To protect users' personal identity information, the system anonymizes and masks direct identifiers (such as names, ID numbers, and mobile phone numbers) before data enters the model for calculation. The model engine processes and analyzes the processed user IDs and their corresponding aggregated behavioral patterns, rather than the original sensitive information traceable to specific individuals.

[0085] Data minimization principle: The data collected by the system is closely related to the purpose of credit assessment and is only necessary for evaluating agricultural management capabilities. For example, it collects the fact that the user "learned cold-weather protection techniques," rather than the biometric information of the user watching course videos; it collects the management result of "600 kg per mu," rather than the specific income and expenditure records of the farmer's household. The system does not collect redundant personal information unrelated to the purpose of credit assessment.

[0086] Closed-loop data sourcing: The multi-dimensional behavioral data relied upon in this invention primarily originates from user business behavior data generated within the agricultural industry chain platform (such as taking courses on the platform, purchasing agricultural supplies in the platform's online store, undertaking agricultural machinery services through the platform, and selling agricultural products on the platform). This constitutes a relatively closed and controllable data ecosystem, with clear and traceable data sources, greatly reducing the compliance risks associated with obtaining data from uncontrollable external channels.

[0087] Security measures: The system employs encrypted transmission, access control, and secure storage technologies to prevent user data from being leaked, tampered with, or illegally accessed during transmission and storage.

[0088] Through the above design, this invention fully utilizes the value of data to improve the accuracy of credit assessment, while protecting user privacy rights from a technical implementation perspective, and ensuring the compliance and feasibility of the technical solution.

[0089] Through the description of specific embodiments, a more in-depth and specific understanding should be gained of the technical means and effects adopted by the present invention to achieve the intended purpose. However, the accompanying drawings are only provided for reference and illustration and are not intended to limit the present invention.

Claims

1. A dynamic credit scoring method based on behavioral data from the entire agricultural industry chain, characterized in that, include: S10, collect multi-dimensional behavioral data of users, including at least: agricultural technology learning and application behavioral data, industrial chain transaction and operation data, and environmental and time series data; S20, Processing the collected multidimensional behavioral data, including: Seasonal weight adaptive: The weights of different evaluation dimensions in the credit scoring model are dynamically adjusted according to the preset agricultural calendar corresponding to the user. Regionality correction: The user's operating indicators are compared with the average level of farmers in the same region and of the same product category in the database to obtain a relative evaluation score; Deep causal analysis: Correlation analysis of users' agricultural input data, agricultural machinery usage data and their final agricultural output data to calculate input efficiency and asset revenue generation capacity; S30 inputs the processed indicator data into the credit scoring model to calculate and output the user's credit score.

2. The dynamic credit scoring method based on behavioral data of the entire agricultural industry chain according to claim 1, characterized in that, The agricultural technology learning and application behavior data in step S10 includes: user viewing and response records of agricultural risk warnings, learning and test completion records of agricultural technology courses, and agricultural input or agricultural machinery consumption records triggered after learning. In step S20, the viewing and response records, learning and quiz completion records, and consumption records are quantitatively evaluated as a continuous behavioral chain.

3. The dynamic credit scoring method based on behavioral data across the entire agricultural industry chain as described in claim 1, characterized in that, The seasonal weight adaptation includes: Identify the current stage of agricultural activity the user is in; The system invokes predefined weight configurations for agricultural stages, where, during the capital investment period, the weight of cash flow-related indicators is reduced, while the weight of indicators related to the preparation of production materials and the application of technology is increased.

4. The dynamic credit scoring method based on behavioral data of the entire agricultural industry chain according to claim 1, characterized in that, The industrial chain transaction and operation data includes: primary industry production data, secondary industry processing and order data, and tertiary industry market and sales data; The method further includes: establishing business relationships between the production data of the first industry, the processing and order data of the second industry, and the market and sales data of the third industry, and assessing the user's operational stability and supply chain resilience based on the relationships.

5. The dynamic credit scoring method based on behavioral data across the entire agricultural industry chain as described in claim 4, characterized in that, Establishing the aforementioned business relationship includes: Identify the degree of matching between primary industry production standards and secondary industry purchase order quality requirements; Identify the compatibility between secondary industry product output and tertiary industry sales channels and prices.

6. The dynamic credit scoring method based on behavioral data across the entire agricultural industry chain as described in claim 4 or 5, characterized in that, The procurement contract documents are automatically identified and structured using optical character recognition technology to collect secondary industry processing and order data.

7. A dynamic credit scoring system based on behavioral data across the entire agricultural industry chain, characterized in that, The dynamic credit scoring system is used to implement the dynamic credit scoring based on agricultural whole-industry chain behavioral data as described in any one of claims 1-6, and the dynamic credit scoring system includes: The data acquisition module is configured to collect multidimensional behavioral data from users. The data processing module is configured to process the multidimensional data, including performing seasonal weight adaptation, regional correction, and deep causal analysis. The model engine module is configured to store and run credit scoring models, using processed indicator data as input to calculate credit scores. The application output module is configured to output credit scores, ratings, and rating reports.

8. The dynamic credit scoring system based on behavioral data across the entire agricultural industry chain as described in claim 7, characterized in that, The dynamic credit scoring system also includes: The scenario-based service encapsulation module is configured to match and provide scenario-based financial services corresponding to specific links in the agricultural industry chain based on the user's credit score.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the dynamic credit scoring method based on behavioral data of the entire agricultural industry chain as described in any one of claims 1 to 6.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the dynamic credit scoring method based on behavioral data across the entire agricultural industry chain as described in any one of claims 1 to 6.

Images