Regional farm product data interaction management method and platform based on multi-source heterogeneous data fusion

Abstract

The invention relates to the technical field of agricultural data management, in particular to a regional farm product data interaction management method and platform based on multi-source heterogeneous data fusion, and the method comprises the steps: obtaining the multi-source data of regional farm products from a plurality of heterogeneous data sources, the multi-source data comprises production data, market data, environment data and user generation data; performing data fusion processing on the multi-source data to generate a fused data set; receiving an interaction request input by a user in real time, executing interaction management analysis based on the fused data set and the interaction request, and generating personalized interaction response data; and generating a user interaction result according to the personalized interaction response data, and feeding back the user interaction result to the user side. The method has the effect of improving the interaction efficiency and decision support capability of the regional farm product data platform.

Description

Technical Field

[0001] This application relates to the technical field of agricultural data management, and in particular to a regional agricultural product data interaction management method and platform based on multi-source heterogeneous data fusion. Background Technology

[0002] With the rapid development of agricultural informatization and digital agriculture, regional agricultural product data platforms have become important tools for improving agricultural production efficiency and market responsiveness. These platforms aim to integrate multi-source data from agricultural production, markets, and the environment to provide data support and decision-making assistance for farmers, distributors, and relevant managers.

[0003] However, existing regional agricultural product data platforms still face numerous challenges in practical applications. Firstly, the data sources are diverse and varied in structure, including production data collected by IoT sensors, price data from market trading systems, environmental data from meteorological departments, and user-entered operation records. These data differ significantly in format, accuracy, and update frequency, resulting in severe data heterogeneity. Traditional platforms often lack effective data fusion mechanisms, leading to widespread data silos and an inability to form a unified and consistent data view, severely limiting the comprehensive utilization value of the data.

[0004] Secondly, existing platforms primarily focus on data collection and storage, offering limited user interaction capabilities. They typically only provide basic data query and display services, lacking intelligent interactive analysis capabilities. When users initiate complex data analysis requests, such as production forecasting or market trend analysis, the platforms often fail to provide accurate predictions based on multi-source data, and struggle to generate personalized responses according to user preferences, resulting in a poor user experience and limiting the platform's practical value.

[0005] Furthermore, existing systems often employ fixed weight allocations and simple statistical algorithms in data processing and analysis, lacking dynamic optimization mechanisms based on data quality and user behavior. This rigid approach struggles to adapt to the complex and ever-changing nature of agricultural data, resulting in insufficient accuracy of analysis results and failing to provide reliable support for agricultural production and management decisions. Simultaneously, platforms generally lack effective feedback and recommendation mechanisms, failing to provide targeted data recommendation services based on users' historical behavior, thus reducing user stickiness and platform efficiency.

[0006] In summary, traditional data management methods struggle to handle the fusion of multi-source, heterogeneous data in the agricultural sector, while simple interactive designs fail to meet the growing personalized needs of users. Therefore, an innovative data interaction management method is urgently needed. Summary of the Invention

[0007] To improve the interactive efficiency and decision support capabilities of regional agricultural product data platforms, this application provides a regional agricultural product data interaction management method and platform based on multi-source heterogeneous data fusion.

[0008] The above-mentioned objective of this application is achieved through the following technical solution:

[0009] A regional agricultural product data interaction management method based on multi-source heterogeneous data fusion includes the following steps:

[0010] Multi-source data on regional agricultural products is obtained from multiple heterogeneous data sources, including production data, market data, environmental data, and user-generated data.

[0011] The multi-source data is fused to generate a fused dataset;

[0012] The system receives user-input interaction requests in real time, performs interaction management analysis based on the fused dataset and interaction requests, and generates personalized interaction response data.

[0013] The user interaction results are generated based on the personalized interaction response data and fed back to the user.

[0014] By adopting the above technical solutions, multi-source data is acquired from multiple heterogeneous data sources, including production, market, environmental, and user-generated data, ensuring the comprehensiveness and diversity of the data. This covers the entire chain of agricultural product information from production to market, providing a rich and real-time input foundation for subsequent analysis. Data fusion processing of the multi-source data generates a fused dataset. Through data cleaning, transformation, and integration, data silos are eliminated, improving data consistency and reliability, and reducing decision-making errors caused by data heterogeneity. Real-time reception of user interaction requests and execution of interaction management analysis, along with dynamic generation of personalized response data based on the fused dataset, enable the platform to quickly adapt to changes in user needs, improving interaction efficiency and user experience. User interaction results are generated and fed back based on the response data, ensuring users receive timely, intuitive, and actionable decision support. This overall improves the interaction efficiency and decision-making accuracy of the regional agricultural product data platform, solving the problems of data dispersion and delayed response in traditional platforms.

[0015] In a preferred embodiment, this application can be further configured such that: the data fusion processing of the multi-source data to generate a fused dataset specifically includes:

[0016] The collected data is cleaned and normalized to remove outliers and missing values, and then standardized to convert the data into a uniform unit of measurement.

[0017] Key features are extracted based on feature extraction methods, including environmental features, production features, market features, and user features.

[0018] The extracted key features are fused based on a weight allocation algorithm to generate a fused dataset.

[0019] By adopting the above technical solutions, the collected data is cleaned and normalized to remove outliers and missing values, and standardized to unify dimensions, effectively improving data quality and consistency, reducing noise interference in analysis, and ensuring the accuracy of subsequent processing. Key features, including environmental, production, market, and user characteristics, are extracted using feature extraction methods. Representative variables are identified through dimensionality reduction and selection techniques, simplifying data complexity and improving analysis efficiency. Key features are fused based on a weighting algorithm, dynamically allocating weights according to feature reliability, ensuring the dominant role of highly reliable features in decision-making, and generating a highly consistent fused dataset. This provides a reliable foundation for interactive analysis, thereby optimizing data management processes, reducing data processing costs, and enhancing the platform's robustness and adaptability.

[0020] In a preferred embodiment, this application can be further configured such that: the real-time reception of user input interaction requests, and the execution of interaction management analysis based on the fused dataset and the interaction requests, include:

[0021] Based on user input instructions, the request type and parameters are determined, relevant data is retrieved from the fused dataset, and the data is filtered and sorted according to user preferences;

[0022] The retrieved data is analyzed using a pre-defined prediction model to predict the yield, market trends, or risk indicators of regional agricultural products and generate prediction results.

[0023] Personalized interactive response data is generated based on prediction results and user preferences.

[0024] By adopting the above technical solutions, the system determines the request type and parameters based on user input instructions, retrieves relevant data from the fused dataset, and filters and sorts the data according to user preferences. This achieves accurate data matching and personalized services, improving retrieval efficiency and user satisfaction. Pre-set prediction models are used to analyze the retrieved data, predicting output, market trends, or risk indicators. By combining historical data and real-time factors with time series or machine learning models, the accuracy and practicality of predictions are improved, providing users with forward-looking decision support. Personalized interactive response data is generated based on prediction results and user preferences. By dynamically matching user needs, customized reports or suggestions are generated, enhancing the relevance and practicality of the interaction. This overall improves the platform's intelligent response capabilities and decision support effects, solving the problems of insufficient analytical capabilities and generalized responses in traditional platforms.

[0025] In a preferred embodiment, this application can be further configured such that, before analyzing the retrieved data using a preset prediction model, the regional agricultural product data interaction management method based on multi-source heterogeneous data fusion further includes:

[0026] Acquire historical regional environmental data, and construct an initial linear prediction model based on the historical regional environmental data and the fused dataset;

[0027] The gradient descent algorithm is used to optimize the model parameters of the initial linear prediction model in order to minimize the prediction error;

[0028] Based on the optimized model parameters, the initial linear prediction model is adjusted to generate a prediction model.

[0029] By adopting the above technical solution, historical regional environmental data is acquired, and an initial linear prediction model is constructed based on historical data and a fused dataset. This ensures the basic accuracy and adaptability of the model, providing a scientific basis for predictive analysis. The gradient descent algorithm is used to optimize the model parameters to minimize prediction errors. Iterative parameter updates improve the model's convergence speed and accuracy, reduce the risk of overfitting, and enhance prediction reliability. The initial model is adjusted based on the optimized parameters to generate the final prediction model. Testing and verification ensure the model's practicality and robustness, thereby improving the platform's accuracy and efficiency in production and market trend prediction, reducing decision-making uncertainty, and providing users with more reliable data support.

[0030] In a preferred embodiment, this application can be further configured as follows: the step of generating user interaction results based on the personalized interaction response data and feeding them back to the user specifically includes:

[0031] The personalized interactive response data is formatted into a user-readable format, including generating user reports or visualizations.

[0032] The response data is displayed through a user interaction interface, and interactive options are provided.

[0033] By adopting the above technical solutions, personalized interactive response data is formatted into a user-readable form, such as generating reports or visual charts, making complex data intuitive and easy to understand, improving user comprehension and ease of operation, and enhancing decision support effectiveness. Displaying response data and providing interactive options, such as zooming, filtering, and downloading, through user interaction interfaces allows users to dynamically adjust views and save information, improving interactive flexibility and user engagement. Overall, through intuitive presentation and flexible interaction, the user experience is optimized, the effective utilization of platform data is promoted, and the problems of traditional platforms' single output format and poor interactivity are solved, thereby improving the platform's usability and user stickiness.

[0034] In a preferred embodiment, this application can be further configured such that: the extracted key features are fused based on the weight allocation algorithm to generate a fused dataset, specifically including:

[0035] Calculate the reliability score for each key feature and assign weights based on the reliability scores;

[0036] The key feature vectors are weighted and averaged using the assigned weights to obtain a fused feature vector. The fused feature vector is then stored in a fused database to form a fused dataset.

[0037] By adopting the above technical solution, the reliability score of each key feature is calculated and weights are assigned according to the score, ensuring that high-reliability features dominate the fusion process, improving the accuracy and credibility of the fused data, and reducing the impact of low-quality data. The assigned weights are used to perform a weighted average of the key feature vectors to obtain the fused feature vector. Mathematical optimization achieves consistent data integration, enhancing the representativeness and analytical value of the data. The fused feature vector is stored in the fusion database to form a fused dataset, facilitating rapid retrieval and updates, optimizing data management efficiency, and thus improving the platform's data processing capabilities and decision support reliability. This solves the problems of uneven data weights and poor fusion effects in traditional methods.

[0038] In a preferred embodiment, this application can be further configured such that, after generating user interaction results based on the personalized interaction response data and feeding them back to the user, the regional agricultural product data interaction management method based on multi-source heterogeneous data fusion further includes:

[0039] Collect user interaction feedback data, including query records and click behavior, and extract it from user interaction logs;

[0040] Based on user interaction feedback data, user preferences are analyzed, and collaborative filtering algorithms are used to calculate user similarity.

[0041] Recommended farm product data is generated based on the user similarity and then fed back to the user.

[0042] By adopting the above technical solutions, and collecting user interaction feedback data, including query records and click behavior, user behavior patterns are extracted from logs, enabling continuous monitoring and analysis of user behavior and providing a data foundation for personalized services. User preferences are analyzed based on user feedback data, collaborative filtering algorithms are used to calculate user similarity, and mathematical modeling is used to accurately identify user interests, enhancing the accuracy and relevance of recommendations. Recommended agricultural product data is generated and fed back based on user similarity, such as recommending related products or best practices, providing proactive services, improving user engagement and satisfaction, thereby optimizing the platform's interactive experience, solving the problems of generalized recommendations and insufficient user participation in traditional platforms, and enhancing the platform's intelligence and user loyalty.

[0043] The second objective of this invention is achieved through the following technical solution:

[0044] A regional agricultural product data interaction management platform based on multi-source heterogeneous data fusion includes:

[0045] The multi-source data acquisition module is used to acquire multi-source data of regional agricultural products from multiple heterogeneous data sources. The multi-source data includes production data, market data, environmental data, and user-generated data.

[0046] The data fusion processing module is used to perform data fusion processing on the multi-source data to generate a fused dataset.

[0047] The user interaction analysis module is used to receive user-input interaction requests in real time, perform interaction management analysis based on the fused dataset and interaction requests, and generate personalized interaction response data.

[0048] The interaction result feedback module is used to generate user interaction results based on the personalized interaction response data and feed them back to the user.

[0049] By adopting the above technical solutions, multi-source data is acquired from multiple heterogeneous data sources, including production, market, environmental, and user-generated data, ensuring the comprehensiveness and diversity of the data. This covers the entire chain of agricultural product information from production to market, providing a rich and real-time input foundation for subsequent analysis. Data fusion processing of the multi-source data generates a fused dataset. Through data cleaning, transformation, and integration, data silos are eliminated, improving data consistency and reliability, and reducing decision-making errors caused by data heterogeneity. Real-time reception of user interaction requests and execution of interaction management analysis, along with dynamic generation of personalized response data based on the fused dataset, enable the platform to quickly adapt to changes in user needs, improving interaction efficiency and user experience. User interaction results are generated and fed back based on the response data, ensuring users receive timely, intuitive, and actionable decision support. This overall improves the interaction efficiency and decision-making accuracy of the regional agricultural product data platform, solving the problems of data dispersion and delayed response in traditional platforms.

[0050] The above-mentioned objective three of this application is achieved through the following technical solution:

[0051] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described regional agricultural product data interaction management method based on multi-source heterogeneous data fusion.

[0052] The fourth objective of this application is achieved through the following technical solution:

[0053] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described method for regional agricultural product data interaction management based on multi-source heterogeneous data fusion.

[0054] In summary, this application includes at least one of the following beneficial technical effects:

[0055] 1. By acquiring multi-source data from multiple heterogeneous data sources, including production, market, environmental, and user-generated data, the platform ensures comprehensiveness and diversity, covering the entire agricultural product supply chain from production to market, providing a rich and real-time input foundation for subsequent analysis. Data fusion processing of multi-source data generates a fused dataset. Through data cleaning, transformation, and integration, data silos are eliminated, improving data consistency and reliability, and reducing decision-making errors caused by data heterogeneity. Real-time reception of user interaction requests and execution of interaction management analysis, along with dynamic generation of personalized response data based on the fused dataset, enables the platform to quickly adapt to changing user needs, improving interaction efficiency and user experience. The platform generates and provides feedback on user interaction results based on the response data, ensuring users receive timely, intuitive, and actionable decision support. This overall improvement enhances the interaction efficiency and decision-making accuracy of the regional agricultural product data platform, solving the problems of data fragmentation and delayed response in traditional platforms.

[0056] 2. By cleaning and normalizing the collected data, outliers and missing values ​​were removed, and standardized methods were used to unify the dimensions, effectively improving data quality and consistency, reducing noise interference in the analysis, and ensuring the accuracy of subsequent processing. Key features, including environmental, production, market, and user features, were extracted based on feature extraction methods. Representative variables were identified through dimensionality reduction and selection techniques, simplifying data complexity and improving analysis efficiency. Key features were fused based on a weight allocation algorithm, and weights were dynamically allocated according to feature reliability, ensuring the dominant role of highly reliable features in decision-making. This generated a highly consistent fused dataset, providing a reliable foundation for interactive analysis, thereby optimizing the data management process, reducing data processing costs, and enhancing the platform's robustness and adaptability.

[0057] 3. By determining the request type and parameters based on user input commands, relevant data is retrieved from the fused dataset, and filtered and sorted according to user preferences, achieving accurate data matching and personalized services, thus improving retrieval efficiency and user satisfaction. Pre-set prediction models are used to analyze the retrieved data, predicting output, market trends, or risk indicators. By combining historical data and real-time factors with time series or machine learning models, the accuracy and practicality of predictions are improved, providing users with forward-looking decision support. Personalized interactive response data is generated based on prediction results and user preferences. By dynamically matching user needs, customized reports or suggestions are generated, enhancing the relevance and practicality of the interaction. This comprehensively improves the platform's intelligent response capabilities and decision support effects, solving the problems of insufficient analytical capabilities and generalized responses in traditional platforms.

[0058] 4. By collecting user interaction feedback data, including query records and click behavior, user behavior patterns are extracted from the logs, enabling continuous monitoring and analysis of user behavior and providing a data foundation for personalized services. User preferences are analyzed based on user feedback data, and collaborative filtering algorithms are used to calculate user similarity. Mathematical modeling accurately identifies user interests, enhancing the accuracy and relevance of recommendations. Recommended agricultural product data is generated and fed back based on user similarity, such as recommending related products or best practices, providing proactive services, improving user engagement and satisfaction, thereby optimizing the platform's interactive experience, solving the problems of generalized recommendations and insufficient user participation in traditional platforms, and enhancing the platform's intelligence and user loyalty. Attached Figure Description

[0059] Figure 1 This is a flowchart illustrating an embodiment of the regional agricultural product data interaction management method based on multi-source heterogeneous data fusion in this application.

[0060] Figure 2 This is a flowchart illustrating the implementation of step S20 in an embodiment of a regional agricultural product data interaction management method based on multi-source heterogeneous data fusion.

[0061] Figure 3 This is a flowchart illustrating the implementation of step S30 in an embodiment of a regional agricultural product data interaction management method based on multi-source heterogeneous data fusion.

[0062] Figure 4 This is another implementation flowchart of an embodiment of a regional agricultural product data interaction management method based on multi-source heterogeneous data fusion;

[0063] Figure 5 This is a flowchart illustrating the implementation of step S40 in the embodiment of the regional agricultural product data interaction management method based on multi-source heterogeneous data fusion in this application;

[0064] Figure 6 This is a flowchart illustrating the implementation of step S23 in the embodiment of the regional agricultural product data interaction management method based on multi-source heterogeneous data fusion in this application;

[0065] Figure 7 This is another implementation flowchart of the regional agricultural product data interaction management method based on multi-source heterogeneous data fusion in this application;

[0066] Figure 8 This is a schematic diagram of an embodiment of the regional agricultural product data interaction management platform based on multi-source heterogeneous data fusion in this application;

[0067] Figure 9 This is a schematic diagram of a computer device according to an embodiment of this application. Detailed Implementation

[0068] The following is in conjunction with the appendix Figure 1-9 This application will be described in further detail.

[0069] In one embodiment, such as Figure 1 As shown, this application discloses a regional agricultural product data interaction management method based on multi-source heterogeneous data fusion, which specifically includes the following steps:

[0070] S10: Obtain multi-source data on regional agricultural products from multiple heterogeneous data sources, including production data, market data, environmental data, and user-generated data.

[0071] In this embodiment, multi-source data refers to a collection of agricultural-related data from different structures and sources; production data refers to information related to the agricultural product production process, such as crop growth status and soil conditions; market data refers to information on the transaction and demand of agricultural products in the market, such as prices and sales volume; environmental data refers to external natural factors affecting agricultural production, such as weather and soil composition; and user-generated data refers to data generated by farmers or platform users through interaction, such as operation records and feedback. These data sources are heterogeneous, including structured data (such as database records) and unstructured data (such as sensor stream data).

[0072] Specifically, a distributed data acquisition system acquires data in real time or periodically from multiple heterogeneous data sources: real-time production data, including soil moisture, temperature, and crop growth status, is obtained from IoT sensors (such as soil moisture sensors and drone images), with sensor data transmitted to the platform's data buffer via a wireless network; historical price and demand data, including daily price fluctuations and regional demand trends, is obtained from market databases (such as agricultural product trading platform APIs), in structured table format; environmental data, including rainfall, sunshine hours, and soil pH, is obtained from meteorological departments or environmental monitoring stations, accessed via a web service interface; and user input data, including farm operation records (such as sowing time and fertilizer application rate) and user feedback (such as ratings and comments), is obtained from user terminals (such as mobile applications or web platforms), uploaded in JSON or XML format. All data undergoes preliminary verification upon acquisition to ensure data integrity and timeliness, providing a foundation for subsequent fusion processing.

[0073] S20: Perform data fusion processing on the multi-source data to generate a fused dataset.

[0074] In this embodiment, data fusion processing refers to the process of integrating multi-source heterogeneous data into a unified and consistent dataset, including data cleaning, data transformation, and data integration. The fused dataset refers to the standardized dataset formed after fusion processing, used to support subsequent analysis and interaction. Data cleaning aims to remove outliers and missing values, data transformation unifies the data into a standardized format and dimension, and data integration generates consistent data through feature fusion.

[0075] Specifically, the process involves data cleaning of multi-source data, using statistical methods (such as the Z-score algorithm) to identify outliers. Outliers are determined based on historical data distribution; for example, for soil moisture data, if a value deviates from the mean by more than 3 standard deviations, it is marked as an outlier and removed. Simultaneously, interpolation methods (such as linear interpolation) are used to fill in missing values. Secondly, the cleaned data undergoes standardization, converting data from different units to a unified dimension. For example, temperature data is converted to degrees Celsius, and price data to standard currency units, using Min-Max standardization or Z-score standardization to ensure data comparability. Then, a data fusion model is used to integrate the standardized data, generating a fused dataset. This model is based on a weighted allocation algorithm, with weights determined by the reliability of the data source. Reliability is calculated using a data quality assessment database, which is generated by comparing the data source output with a benchmark (such as an official agricultural report). Finally, the fused dataset is stored in a fused database for subsequent interactive management and analysis.

[0076] S30: Receive user input interaction requests in real time, perform interaction management analysis based on the fused dataset and interaction requests, and generate personalized interaction response data.

[0077] In this embodiment, an interaction request refers to an operation instruction initiated by a user through a platform interface, including data query, data sharing, or data analysis requests; interaction management analysis refers to the process of data retrieval, prediction, or optimization analysis based on user requests and integrated data; personalized interaction response data refers to customized results generated based on user preferences and requests, such as reports, charts, or recommendations.

[0078] Specifically, the system receives user-input interaction requests through a user interface, providing a graphical user interface that displays a summary of the fused data and interactive options, including data visualization charts (such as production trend charts) and input forms (such as query condition forms). Secondly, it receives user-selected data query conditions, including product type (such as wheat, corn), time range (such as the most recent month), and regional range (such as a specific farm area), which are used to filter the fused data set. Then, it receives user-uploaded shared data, including farm photos and production reports, and integrates this shared data into the multi-source data, ensuring format compatibility through a data validation module. Finally, it receives user-triggered data analysis requests, including price forecasts and production suggestions, generated based on user input parameters. Based on the interaction requests and the fused dataset, it performs interactive management analysis: parsing the interaction requests to determine the request type and parameters, for example, identifying the query type as "price forecast" or an analysis model requirement; retrieving relevant data from the fused data set, using indexing and query optimization techniques (such as B-tree indexes) to improve retrieval efficiency; and applying analysis models to generate response data according to the request type, for example, returning retrieved data for data query requests and performing prediction or optimization analysis for data analysis requests.

[0079] S40: Generate user interaction results based on the personalized interaction response data and feed them back to the user.

[0080] In this embodiment, user interaction results refer to the data output finally presented to the user, including visual charts, reports, or recommendation lists; feedback to the user terminal refers to returning the results to the user's device through the platform interface.

[0081] Specifically, personalized interactive response data is formatted into a user-readable format, including generating user reports or visualizations, automatically generating report content using a template engine, and displaying response data through a user interaction interface.

[0082] In this embodiment, multi-source data, including production, market, environmental, and user-generated data, is acquired from multiple heterogeneous data sources, ensuring the comprehensiveness and diversity of the data. This covers the entire chain of agricultural product information from production to market, providing a rich and real-time input foundation for subsequent analysis. Data fusion processing is performed on the multi-source data to generate a fused dataset. Data cleaning, transformation, and integration eliminate data silos, improve data consistency and reliability, and reduce decision-making errors caused by data heterogeneity. User interaction requests are received in real time, and interaction management analysis is performed. Personalized response data is dynamically generated based on the fused dataset, enabling the platform to quickly adapt to changes in user needs and improving interaction efficiency and user experience. User interaction results are generated and fed back based on the response data, ensuring that users receive timely, intuitive, and actionable decision support. This overall improves the interaction efficiency and decision-making accuracy of the regional agricultural product data platform, solving the problems of data dispersion and delayed response in traditional platforms.

[0083] In one embodiment, such as Figure 2 As shown, in step S20, the multi-source data is fused to generate a fused dataset, which specifically includes:

[0084] S21: Clean and normalize the collected data, remove outliers and missing values, and use standardization methods to convert the data into a uniform dimension.

[0085] In this embodiment, data cleaning refers to identifying and correcting errors or inconsistencies in the data; normalization refers to scaling the data to a specific range; and standardization methods refer to techniques for converting data into a uniform dimension.

[0086] Specifically, data preprocessing is used to clean the collected data and identify outliers. For example, for production data, if a value exceeds twice the standard deviation of the historical average, it is considered an outlier and removed. At the same time, the mean or median is used to fill missing values. Secondly, normalization is performed to transform the data to the range of [0,1].

[0087] S22: Extract key features based on feature extraction methods, including environmental features, production features, market features, and user features.

[0088] In this embodiment, key features refer to representative attributes extracted from the data, used to simplify data and improve analysis efficiency; environmental features include rainfall, temperature, etc.; production features include output, growth cycle, etc.; market features include price, demand, etc.; user features include user behavior, preferences, etc.

[0089] Specifically, feature extraction algorithms (such as Principal Component Analysis (PCA) or Random Forest feature importance analysis) are applied to extract key features from the cleaned data: for environmental features, rainfall, sunshine duration, and soil composition are extracted from environmental data; for production features, crop growth status, yield, and fertilizer application are extracted from production data; for market features, price fluctuations, demand index, and inventory levels are extracted from market data; and for user features, user activity, feedback ratings, and operation frequency are extracted from user-generated data. The extracted features are stored as feature vectors for subsequent fusion.

[0090] S23: Based on the weight allocation algorithm, the extracted key features are fused to generate a fused dataset.

[0091] In this embodiment, the weight allocation algorithm refers to a mathematical method that assigns weights based on the reliability or importance of features; the fused dataset refers to a unified data set formed by weighted fusion.

[0092] Specifically, a reliability score is calculated for each key feature. This score is determined based on the historical accuracy and feature consistency of the data source. For example, the error rate is calculated by comparing feature values ​​with benchmark data; the score is inversely proportional to the error rate. Next, weights are assigned based on the reliability scores, using linear allocation or optimization algorithms to determine the weight values. Then, the assigned weights are used to perform a weighted average on the key feature vectors to obtain a fused feature vector. For example, for environmental feature vectors, the fused value is calculated as a weighted sum. Finally, the fused feature vector is stored in a fused database to form a fused dataset. The weight allocation formula is as follows:

[0093] ,in, It is the weight of feature i. Let be the reliability score of feature i, and n be the total number of features. The fused feature vector is calculated as follows:

[0094] ,in, It is the vector value of feature i.

[0095] In one embodiment, such as Figure 3 As shown, in step S30, user input interaction requests are received in real time, and interaction management analysis is performed based on the fused dataset and the interaction requests, specifically including:

[0096] S31: Based on user input instructions, determine the request type and parameters, retrieve relevant data from the fused dataset, and filter and sort the data according to user preferences.

[0097] In this embodiment, the request type refers to the type of user operation, such as query, share, or analysis; the parameters refer to the specific conditions of the request, such as time range or product type; and the user preferences refer to the user's historical behavior or settings, used for personalized filtering.

[0098] Specifically, the system parses user input commands through a natural language processing module or form to determine the request type (e.g., "data query" or "price prediction") and parameters (e.g., product type = "wheat", time range = "2023-2024"). Then, it retrieves relevant data from the fused dataset, using SQL queries or a search engine to filter data based on parameters, for example, retrieving agricultural product data for specific regions and times. Finally, it filters and sorts data according to user preferences, extracted from historical user interaction logs. For example, if a user frequently queries organic agricultural products, organic-related data is displayed first, and the sorting is based on relevance scores or timestamps.

[0099] S32: Analyze the retrieved data using a preset prediction model to predict the yield, market trends, or risk indicators of regional agricultural products and generate prediction results.

[0100] In this embodiment, the prediction model refers to a machine learning or statistical model used for predicting future values, such as a time series model or a regression model; the prediction result refers to the estimated value output by the model, such as production forecast or price trend.

[0101] Specifically, the retrieved data is analyzed using pre-defined prediction models: for production forecasting, time series models (such as ARIMA) or machine learning models (such as random forests) are used, with input data including historical production, environmental factors, and production data, and outputting future production estimates; for market trend forecasting, regression analysis or neural networks are used, with input data including historical prices, demand data, and macroeconomic indicators, and outputting price change trends; for risk indicators, classification models (such as logistic regression) are used to assess production risks, with input data including weather anomalies and market fluctuations, and outputting risk levels.

[0102] S33: Generate personalized interactive response data based on prediction results and user preferences.

[0103] In this embodiment, personalized interactive response data refers to customized outputs that combine prediction results and user preferences, such as suggestion reports or alerts.

[0104] Specifically, the prediction results are matched with user preferences. For example, if users prefer high-risk, high-return agricultural products, relevant predictions are highlighted. Then, personalized interactive response data is generated, including text summaries, charts, or action suggestions, and the output is formatted using a template engine. Finally, a user feedback mechanism is integrated, allowing users to rate or adjust the response content to continuously optimize the personalization algorithm.

[0105] In one embodiment, such as Figure 4 As shown, before step S32, that is, before analyzing the retrieved data using a preset prediction model, the regional agricultural product data interaction management method based on multi-source heterogeneous data fusion further includes:

[0106] S301: Obtain historical regional environmental data, and construct an initial linear prediction model based on the historical regional environmental data and the fused dataset.

[0107] In this embodiment, historical regional environmental data refers to past environmental records, such as meteorological data; the initial linear prediction model refers to a simple prediction model based on linear assumptions, such as multiple linear regression.

[0108] Specifically, historical regional environmental data, including rainfall and temperature data from the past five years, is obtained from the fused database. An initial linear prediction model is then constructed based on the historical environmental data and the fused dataset; for example, multiple linear regression is used, where the dependent variable is yield and the independent variables are environmental factors.

[0109]

[0110] Where Y is yield, X1 is rainfall, X2 is temperature, β is the model parameter, and ϵ is the error term. The model parameters are initially estimated using the least squares method.

[0111] S302: Optimize the model parameters of the initial linear prediction model using the gradient descent algorithm to minimize the prediction error.

[0112] S303: Based on the optimized model parameters, adjust the initial linear prediction model to generate a prediction model.

[0113] In this embodiment, gradient descent is an iterative optimization algorithm used to minimize the loss function; prediction error refers to the difference between the model output and the true value.

[0114] Specifically, during training, the gradient of the loss function with respect to each model parameter is calculated, and the parameters are iteratively updated along the reverse direction of the gradient until the loss function converges to a predetermined threshold, thereby obtaining a set of optimized parameters that minimize the prediction error. The optimized parameters are then substituted into the initial linear prediction model to form the final prediction model, which is then deployed to the platform analysis module for real-time prediction.

[0115] In one embodiment, such as Figure 5 As shown, in step S40, the user interaction result is generated based on the personalized interaction response data and fed back to the user, specifically including:

[0116] S41: Format the personalized interactive response data into a user-readable format, including generating user reports or visualizations.

[0117] In this embodiment, user-readable format refers to an output format that is easy to understand, such as a natural language report or a graph; visual charts refer to a graphical representation of data, such as a line chart or a pie chart.

[0118] Specifically, report generation tools are used to convert personalized interactive response data into user reports, including text summaries and key metrics; meanwhile, visualization libraries are used to generate charts, such as production trend line charts or market heat maps; reports and charts are integrated into a web interface to support interactive operations.

[0119] S42: Displays response data through a user interaction interface and provides interactive options.

[0120] Specifically, the user interface refers to the interface through which the platform interacts with the user, such as a web page or mobile application; interaction options refer to the actions that the user can perform, such as filtering and downloading. This step enhances interaction flexibility by displaying responsive data through responsive web design, ensuring cross-device compatibility; it provides interaction options such as data filtering (by time or product), chart zooming, and data export; user actions are responded to in real time through the event handling module, updating the displayed content.

[0121] In one embodiment, such as Figure 6 As shown, in step S23, the extracted key features are fused based on the weight allocation algorithm to generate a fused dataset, which specifically includes:

[0122] S231: Calculate the reliability score for each key feature and assign weights based on the reliability scores.

[0123] In this embodiment, the reliability score refers to a quantitative indicator of the quality of the feature data; the weight allocation refers to determining the importance of the feature in the fusion based on the score.

[0124] Specifically, the reliability score is calculated based on data source accuracy, data freshness, and consistency. For example, the scoring formula is:

[0125] ;

[0126] Here, accuracy is the accuracy rate (compared to benchmark data), freshness is the data update time, consistency is the consistency with other data sources, and α, β, γ are weight coefficients that are set according to domain knowledge. Then, the scores are used to assign weights, and the weights are proportional to the scores.

[0127] S232: Use the assigned weights to perform a weighted average on the key feature vectors to obtain a fused feature vector, and store the fused feature vector in the fused database to form a fused dataset.

[0128] Specifically, weighted average refers to calculating the average value according to weights; fused feature vector refers to the fused feature representation, and a weighted average is calculated for each feature vector. For example, for environmental feature vectors, the fused value is calculated as the weighted sum; then the fused feature vector is stored in the fused database, and index optimization is used for fast retrieval.

[0129] In one embodiment, such as Figure 7 As shown, after step S40, that is, after generating user interaction results based on the personalized interaction response data and feeding them back to the user, the regional agricultural product data interaction management method based on multi-source heterogeneous data fusion further includes:

[0130] S50: Collect user interaction feedback data, including query records and click behavior, and extract it from user interaction logs.

[0131] In this embodiment, user interaction feedback data refers to behavioral data generated by user interaction with the platform; user interaction log refers to log files that record user operations.

[0132] Specifically, user interaction feedback data is collected from the platform's log system, including query keywords, number of clicks, and dwell time; the data is stored in a structured format for analyzing user behavior patterns.

[0133] S60: Analyze user preferences based on user interaction feedback data and use collaborative filtering algorithms to calculate user similarity.

[0134] In this embodiment, user preferences refer to users' interests; collaborative filtering is a recommendation algorithm that generates recommendations based on user similarity.

[0135] Specifically, the process involves analyzing user interaction feedback data to extract user preferences, for example, by identifying user groups through clustering algorithms; then, collaborative filtering algorithms are used to calculate user similarity based on the user-item rating matrix, with the cosine similarity formula used for the similarity calculation; finally, a recommendation list is generated based on the behavior of similar users.

[0136] S70: Generate recommended agricultural product data based on the user similarity and feed it back to the user.

[0137] Specifically, recommended agricultural product data refers to agricultural product information tailored to user interests. Based on user similarity, recommended agricultural product data is generated for target users, such as recommending highly rated products purchased by similar users. Then, the recommended data is integrated into the user interaction results and fed back to the user through push notifications or interface display.

[0138] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0139] In one embodiment, a regional agricultural product data interaction management platform based on multi-source heterogeneous data fusion is provided. This platform corresponds one-to-one with the regional agricultural product data interaction management method based on multi-source heterogeneous data fusion described in the above embodiments. Figure 8 As shown, this regional agricultural product data interaction management platform based on multi-source heterogeneous data fusion includes:

[0140] The multi-source data acquisition module is used to acquire multi-source data of regional agricultural products from multiple heterogeneous data sources. The multi-source data includes production data, market data, environmental data, and user-generated data.

[0141] The data fusion processing module is used to perform data fusion processing on the multi-source data to generate a fused dataset.

[0142] The user interaction analysis module is used to receive user-input interaction requests in real time, perform interaction management analysis based on the fused dataset and interaction requests, and generate personalized interaction response data.

[0143] The interaction result feedback module is used to generate user interaction results based on the personalized interaction response data and feed them back to the user.

[0144] Specific limitations regarding the regional agricultural product data interaction management platform based on multi-source heterogeneous data fusion can be found in the limitations of the regional agricultural product data interaction management method based on multi-source heterogeneous data fusion mentioned above, and will not be repeated here. Each module in the aforementioned regional agricultural product data interaction management platform based on multi-source heterogeneous data fusion can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0145] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 9As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database stores data. The network interface communicates with external terminals via a network connection. When executed by the processor, the computer program implements a regional agricultural product data interaction management method based on multi-source heterogeneous data fusion.

[0146] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps:

[0147] Multi-source data on regional agricultural products is obtained from multiple heterogeneous data sources, including production data, market data, environmental data, and user-generated data.

[0148] The multi-source data is fused to generate a fused dataset;

[0149] The system receives user-input interaction requests in real time, performs interaction management analysis based on the fused dataset and interaction requests, and generates personalized interaction response data.

[0150] The user interaction results are generated based on the personalized interaction response data and fed back to the user.

[0151] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0152] Multi-source data on regional agricultural products is obtained from multiple heterogeneous data sources, including production data, market data, environmental data, and user-generated data.

[0153] The multi-source data is fused to generate a fused dataset;

[0154] The system receives user-input interaction requests in real time, performs interaction management analysis based on the fused dataset and interaction requests, and generates personalized interaction response data.

[0155] The user interaction results are generated based on the personalized interaction response data and fed back to the user.

[0156] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0157] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0158] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A regional agricultural product data interaction management method based on multi-source heterogeneous data fusion, characterized in that, The regional agricultural product data interaction management method based on multi-source heterogeneous data fusion includes the following steps: Multi-source data on regional agricultural products is obtained from multiple heterogeneous data sources, including production data, market data, environmental data, and user-generated data. The multi-source data is fused to generate a fused dataset; The system receives user-input interaction requests in real time, performs interaction management analysis based on the fused dataset and interaction requests, and generates personalized interaction response data. The user interaction results are generated based on the personalized interaction response data and fed back to the user.

2. The regional agricultural product data interaction management method based on multi-source heterogeneous data fusion according to claim 1, characterized in that: The step of performing data fusion processing on the multi-source data to generate a fused dataset specifically includes: The collected data is cleaned and normalized to remove outliers and missing values, and then standardized to convert the data into a uniform unit of measurement. Key features are extracted based on feature extraction methods, including environmental features, production features, market features, and user features. The extracted key features are fused based on a weight allocation algorithm to generate a fused dataset.

3. The regional agricultural product data interaction management method based on multi-source heterogeneous data fusion according to claim 1, characterized in that: The real-time reception of user input interaction requests, and the execution of interaction management analysis based on the fused dataset and interaction requests, specifically include: Based on user input instructions, the request type and parameters are determined, relevant data is retrieved from the fused dataset, and the data is filtered and sorted according to user preferences; The retrieved data is analyzed using a pre-defined prediction model to predict the yield, market trends, or risk indicators of regional agricultural products and generate prediction results. Personalized interactive response data is generated based on prediction results and user preferences.

4. The regional agricultural product data interaction management method based on multi-source heterogeneous data fusion according to claim 3, characterized in that: Before analyzing the retrieved data using a preset prediction model, the regional agricultural product data interaction management method based on multi-source heterogeneous data fusion further includes: Acquire historical regional environmental data, and construct an initial linear prediction model based on the historical regional environmental data and the fused dataset; The gradient descent algorithm is used to optimize the model parameters of the initial linear prediction model in order to minimize the prediction error; Based on the optimized model parameters, the initial linear prediction model is adjusted to generate a prediction model.

5. The regional agricultural product data interaction management method based on multi-source heterogeneous data fusion according to claim 1, characterized in that: The step of generating user interaction results based on the personalized interaction response data and feeding them back to the user specifically includes: The personalized interactive response data is formatted into a user-readable format, including generating user reports or visualizations. The response data is displayed through a user interaction interface, and interactive options are provided.

6. The regional agricultural product data interaction management method based on multi-source heterogeneous data fusion according to claim 2, characterized in that: The weighted allocation algorithm is used to fuse the extracted key features to generate a fused dataset, specifically including: Calculate the reliability score for each key feature and assign weights based on the reliability scores; The key feature vectors are weighted and averaged using the assigned weights to obtain a fused feature vector. The fused feature vector is then stored in a fused database to form a fused dataset.

7. The regional agricultural product data interaction management method based on multi-source heterogeneous data fusion according to claim 1, characterized in that: After generating user interaction results based on the personalized interaction response data and feeding them back to the user, the regional agricultural product data interaction management method based on multi-source heterogeneous data fusion further includes: Collect user interaction feedback data, including query records and click behavior, and extract it from user interaction logs; Based on user interaction feedback data, user preferences are analyzed, and collaborative filtering algorithms are used to calculate user similarity. Recommended farm product data is generated based on the user similarity and then fed back to the user.

8. A regional agricultural product data interaction management platform based on multi-source heterogeneous data fusion, characterized in that, include: The multi-source data acquisition module is used to acquire multi-source data of regional agricultural products from multiple heterogeneous data sources. The multi-source data includes production data, market data, environmental data, and user-generated data. The data fusion processing module is used to perform data fusion processing on the multi-source data to generate a fused dataset. The user interaction analysis module is used to receive user-input interaction requests in real time, perform interaction management analysis based on the fused dataset and interaction requests, and generate personalized interaction response data. The interaction result feedback module is used to generate user interaction results based on the personalized interaction response data and feed them back to the user.

9. A computer 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 computer program, it implements the steps of the regional agricultural product data interaction management method based on multi-source heterogeneous data fusion as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the regional agricultural product data interaction management method based on multi-source heterogeneous data fusion as described in any one of claims 1 to 7.

Images