Food safety data operation and maintenance analysis method and system based on big data

Abstract

The invention discloses a food safety data operation and maintenance analysis method and system based on big data. The intelligent food safety information tracing method comprises the following steps: S110, collecting food safety related data as input data; s120, generating a food safety feature code for the input data; s130, quantifying the possibility of food safety risks in the data; s140, calculating a risk fusion score according to the food safety feature code and the risk possibility; s150, performing food safety risk early warning and response according to the risk fusion score; s160, continuously optimizing the food safety data model; according to the method, continuous data acquisition and analysis are provided for food safety information operation and maintenance analysis, and the risk model is dynamically optimized, so that closed-loop operation of risk management is realized, and the overall prevention and control capability and response efficiency of the system are improved.

Description

Technical Field

[0001] The present invention relates to big data technology, and in particular to a food safety data operation and maintenance analysis method based on big data. Background Art

[0002] Intelligent food safety information traceability methods aim to achieve traceability and safety assurance throughout the entire food supply chain, from production to consumption, through advanced technologies. With the growing global food safety crisis, traditional food traceability methods are no longer sufficient to meet the demands of real-time monitoring and efficient management. Therefore, leveraging technologies such as the Internet of Things (IoT), blockchain, data fusion, and artificial intelligence, intelligent food safety traceability systems are being developed. These systems can collect and monitor data from food production, processing, transportation, storage, and other stages in real time. Through multi-source data fusion, dynamic risk assessment, and path optimization, these systems can enhance food safety management. Furthermore, intelligent early warning systems can quickly identify potential risks and automatically initiate emergency response measures, minimizing food safety risks and ensuring transparency and traceability throughout the food supply chain. This approach not only improves food safety assurance capabilities but also enhances consumer trust and satisfaction with food quality. Summary of the Invention

[0003] The present invention provides a food safety data operation and maintenance analysis method based on big data, comprising:

[0004] S110. Collect food safety related data as input data;

[0005] S120, generating a food safety characteristic code for the input data;

[0006] S130, Quantify the likelihood of food safety risks in data;

[0007] S140. Calculate the risk fusion score based on the food safety characteristic code and risk probability;

[0008] S150. Conduct food safety risk warning and response based on risk integration scores;

[0009] S160. Continuously optimize food safety data models.

[0010] As described above, a food safety data operation and maintenance analysis method based on big data is described, in which food safety-related data is collected as input data, and multi-source data collection tools are deployed, including environmental sensor networks, laboratory testing systems, logistics monitoring equipment, and public opinion analysis platforms, to automatically collect various data sources. Production environment data is transmitted to the data center in real time through Internet of Things devices, product testing data is uploaded through laboratory information systems or rapid testing equipment, supply chain logistics data is recorded and transmitted through GPS and temperature and humidity sensors, and consumer feedback data uses natural language processing technology to extract key information from text.

[0011] The method for generating food safety feature codes in the food safety data operation and maintenance analysis method based on big data described above specifically includes the following sub-steps:

[0012] Extract local features from input data and perform weighting and normalization;

[0013] Map the fused features to the risk interval through the activation function;

[0014] Generate food safety signature codes for processed input data.

[0015] As described above, a food safety data operation and maintenance analysis method based on big data is proposed, in which the calculation model outputs a risk gradient with respect to the input. The risk gradient reflects the model's sensitivity to risks near the current input. In order to distinguish normal fluctuations from real risks, an adaptive adjustment factor based on the distance between the input and historical safety data is introduced. This factor tends to zero when the input is close to the safety data, suppressing false positives, and tends to one when the input deviates from the safety data, exposing potential risks.

[0016] The above-mentioned food safety data operation and maintenance analysis method based on big data, wherein the method of calculating the risk fusion score based on the food safety feature coding and risk probability specifically includes the following sub-steps:

[0017] Calculate feature encoding bias;

[0018] Calculate the risk fusion score based on feature coding deviation and risk probability.

[0019] The present invention also provides a food safety data operation and maintenance analysis system based on big data, including: a data acquisition module, a food safety feature code generation module, a food safety risk possibility module, a risk fusion score calculation module, an early warning and response module and a continuous optimization module.

[0020] Data collection module: collects food safety related data as input data;

[0021] Food safety feature code generation module: generates food safety feature codes for input data;

[0022] Food Safety Risk Likelihood Module: Quantifies the likelihood of food safety risks in data;

[0023] Risk fusion score calculation module: calculates the risk fusion score based on food safety feature coding and risk probability;

[0024] Early warning and response module: Provides food safety risk early warning and response based on risk integration scores;

[0025] Continuous optimization module: Continuously optimize the food safety data model.

[0026] As described above, a food safety data operation and maintenance analysis system based on big data collects food safety-related data as input data, deploys multi-source data collection tools, including environmental sensor networks, laboratory testing systems, logistics monitoring equipment, and public opinion analysis platforms, to automatically collect various data sources. Production environment data is transmitted to the data center in real time through IoT devices, product testing data is uploaded through laboratory information systems or rapid testing equipment, supply chain logistics data is recorded and transmitted through GPS and temperature and humidity sensors, and consumer feedback data uses natural language processing technology to extract key information from text.

[0027] In the food safety data operation and maintenance analysis system based on big data, the method for generating food safety feature codes specifically includes the following sub-steps:

[0028] Extract local features from input data and perform weighting and normalization;

[0029] Map the fused features to the risk interval through the activation function;

[0030] Generate food safety signature codes for processed input data.

[0031] As described above, a food safety data operation and maintenance analysis system based on big data is described, in which the computational model outputs a risk gradient about the input. The risk gradient reflects the model's sensitivity to risks near the current input. In order to distinguish normal fluctuations from real risks, an adaptive adjustment factor based on the distance between the input and historical safety data is introduced. This factor tends to zero when the input is close to the safety data, suppressing false positives, and tends to one when the input deviates from the safety data, exposing potential risks.

[0032] In the aforementioned food safety data operation and maintenance analysis system based on big data, the method for calculating the risk fusion score based on the food safety feature code and risk likelihood specifically includes the following sub-steps:

[0033] Calculate feature encoding bias;

[0034] Calculate the risk fusion score based on feature coding deviation and risk probability.

[0035] The beneficial effects achieved by the present invention are as follows: This application provides continuous data collection and analysis for food safety information operation and maintenance analysis, dynamically optimizes risk models, thereby realizing closed-loop operation of risk management and improving the overall prevention and control capabilities and response efficiency of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments described in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0037] Figure 1 This is a flow chart of a food safety data operation and maintenance analysis method based on big data provided in Example 1 of the present application;

[0038] Figure 2 This is a schematic diagram of a food safety data operation and maintenance analysis system based on big data provided in Example 2 of the present application. DETAILED DESCRIPTION

[0039] The following is a clear and complete description of the technical solutions in the embodiments of the present invention, in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0040] Example 1

[0041] like Figure 1 As shown, the first embodiment of the present application provides a food safety data operation and maintenance analysis method based on big data

[0042] Step S110: collecting food safety related data as input data;

[0043] Sources of food safety data primarily include production environment data, product testing data, supply chain logistics data, and consumer feedback data. Production environment data refers to environmental parameters such as temperature, humidity, air quality, and water quality collected through sensors or IoT devices. These data are used to assess the sanitary conditions of the food production environment. Product testing data refers to indicators such as microbial content, pesticide residues, and additives obtained by laboratories or rapid testing equipment. These data directly reflect food safety. Supply chain logistics data includes transportation temperature, storage conditions, and transportation time. These data help track potential risks in the food supply chain. Consumer feedback data refers to consumer health feedback or complaint information collected through channels such as social media and complaint platforms. These data can be used to detect early signs of food safety issues.

[0044] To collect food safety-related data as input, we first deploy multi-source data collection tools, including environmental sensor networks, laboratory testing systems, logistics monitoring equipment, and public opinion analysis platforms, to automate the collection of various data sources. Production environment data is transmitted to the data center in real time via IoT devices. Product testing data is uploaded via laboratory information systems or rapid testing equipment. Supply chain logistics data is recorded and transmitted using GPS and temperature and humidity sensors. Consumer feedback data is extracted from text using natural language processing techniques. All data is cleaned and standardized, stored in a distributed database to ensure data integrity and traceability, and then passed as input to the food safety analysis system.

[0045] Step S120: Generate a food safety feature code for the input data;

[0046] A food safety signature code represents a unique, robust code generated for each piece of input data. It is used to identify and distinguish different data samples and track potential food safety risks. The signature code of the input data is calculated using a feature extraction network. If anomalies or risks exist in the data, the signature code will significantly deviate from the normal range. The method for generating a food safety signature code specifically includes the following sub-steps:

[0047] Step S121: extract local features from the input data, and perform weighting and normalization processing;

[0048] The input data is divided into four dimensions: production environment, product testing, supply chain logistics, and consumer feedback. Each dimension is assigned a learnable weight and a weighted fusion value is calculated to highlight key features. Specifically, the formula is used: Perform weighted fusion processing on the input data, where F represents the weighted fusion result, f i represents the eigenvalue of the i-th dimension, ω i represents the learnable weight of the i-th dimension.

[0049] Step S122: Mapping the fused features to the risk interval through an activation function;

[0050] The weighted fused features are mapped to the risk interval through the activation function to enhance the sensitivity to outliers. The activation function adopts the ReLU function. Specifically, the formula: R = ReLU (F-τ) is used to represent the mapping process, where R represents the risk feature after mapping, τ represents the risk threshold, F represents the weighted fusion result, and ReLU represents the activation function.

[0051] Step S123: Generate a food safety feature code for the processed input data;

[0052] The multi-dimensional correlation of data is utilized to make the coding specific to different risks. Specifically, the formula: C = R × M is used to generate food safety feature coding, where C represents the generated food safety feature coding, R represents the mapped risk feature, M represents the sensitive adjustment matrix related to food safety, and × represents matrix multiplication.

[0053] Step S130: quantifying the possibility of food safety risks in the data;

[0054] To quantify the likelihood of food safety risks in input data, a risk gradient field is introduced. This gradient field can quantify the direction and magnitude of data anomalies in the feature space. First, the risk gradient of the model output with respect to the input is calculated. The risk gradient reflects the model's sensitivity to risks near the current input. To distinguish normal fluctuations from real risks, an adaptive adjustment factor based on the distance between the input and historical safety data is introduced. This factor tends to zero when the input is close to the safe data, suppressing false positives; and tends to one when the input deviates from the safe data, exposing potential risks. Specifically, the formula is used: Calculate the possibility of food safety risk, where P represents the possibility of food safety risk, x represents the input data, and x represents the input data. safe represents historical safety data, p(y|x) represents the prediction model, and λ represents the adjustment coefficient.

[0055] Step S140: Calculate a risk fusion score based on the food safety feature code and risk likelihood;

[0056] The method for calculating the risk fusion score based on the food safety characteristic code and risk probability includes the following sub-steps:

[0057] Step S141, calculating feature coding deviation;

[0058] In order to reflect the degree of deviation between the input sample and the safe sample in the feature space, the food safety feature code C generated by the input data x is compared with the historical safe sample x. safeThe generated security feature code is compared and the deviation value is calculated. Specifically, the formula is: D = || CC safe ||2 Calculate the feature coding deviation, where D represents the deviation value, C represents the food safety feature code, and C safe Represents the security feature code generated by historical security samples.

[0059] Step S142: Calculate the risk fusion score based on the feature coding deviation and risk probability;

[0060] The risk fusion score represents a quantitative indicator of food safety risk. The larger the value, the higher the risk. Specifically, the risk fusion score is calculated using the formula: S = α·D + β·P, where S represents the risk fusion score, and α and β represent the influencing factors of feature coding deviation D and risk probability P, respectively.

[0061] Step S150: Conduct food safety risk warning and response based on the risk fusion score;

[0062] When the risk fusion score S exceeds the set safety threshold, the system automatically triggers an early warning mechanism to promptly respond to potential risks. If the high score is due to feature encoding deviations, the system immediately initiates anomaly detection processes within the production environment or supply chain to locate and isolate potentially problematic batches of products to prevent further problems. Simultaneously, the system conducts in-depth analysis of the corresponding data to identify the specific links causing the scoring anomaly and implements measures to correct the model deviations to ensure the accuracy and reliability of the scoring.

[0063] If a high score reflects the possibility of an actual risk, the system will quickly initiate laboratory re-inspections or leverage consumer feedback to verify the true source of the risk, ensuring that the risk assessment is scientifically based. At the same time, the system will automatically generate a detailed risk report, send an early warning notification to the relevant responsible departments, and initiate corresponding emergency plans based on the risk level, such as recalling problematic products, adjusting production processes, or strengthening quality control. Throughout this process, the system will continuously collect and analyze new data and dynamically optimize the risk model, thereby achieving a closed-loop risk management system and improving the system's overall prevention and control capabilities and response efficiency.

[0064] Step S160: Continuously optimize the food safety data model;

[0065] By deeply mining historical data and continuously collecting real-time feedback, the system regularly updates its feature extraction network and risk quantification model, thereby continuously improving its ability to identify new risks. This process not only helps discover potential risk characteristics but also strengthens the model's adaptability and judgment in different scenarios, making risk identification more intelligent and precise, and ensuring the overall security of products and services.

[0066] The system also dynamically adjusts security thresholds and risk response strategies based on the latest industry standards, regulatory requirements, and changes in laws and policies. This mechanism allows the system to automatically adapt to changes in the external environment at different time points, ensuring the timeliness and foresight of risk prevention and control measures. While ensuring compliance, the risk management system can also continuously improve its accuracy and responsiveness, achieving comprehensive and efficient risk governance.

[0067] Example 2

[0068] like Figure 2 As shown, the second embodiment of the present application provides a food safety data operation and maintenance analysis system based on big data, including:

[0069] Data collection module 21: collects food safety related data as input data;

[0070] Sources of food safety data primarily include production environment data, product testing data, supply chain logistics data, and consumer feedback data. Production environment data refers to environmental parameters such as temperature, humidity, air quality, and water quality collected through sensors or IoT devices. These data are used to assess the sanitary conditions of the food production environment. Product testing data refers to indicators such as microbial content, pesticide residues, and additives obtained by laboratories or rapid testing equipment. These data directly reflect food safety. Supply chain logistics data includes transportation temperature, storage conditions, and transportation time. These data help track potential risks in the food supply chain. Consumer feedback data refers to consumer health feedback or complaint information collected through channels such as social media and complaint platforms. These data can be used to detect early signs of food safety issues.

[0071] To collect food safety-related data as input, we first deploy multi-source data collection tools, including environmental sensor networks, laboratory testing systems, logistics monitoring equipment, and public opinion analysis platforms, to automate the collection of various data sources. Production environment data is transmitted to the data center in real time via IoT devices. Product testing data is uploaded via laboratory information systems or rapid testing equipment. Supply chain logistics data is recorded and transmitted using GPS and temperature and humidity sensors. Consumer feedback data is extracted from text using natural language processing techniques. All data is cleaned and standardized, stored in a distributed database to ensure data integrity and traceability, and then passed as input to the food safety analysis system.

[0072] Food safety feature code generation module 22: generates food safety feature code for input data;

[0073] A food safety signature code represents a unique, robust code generated for each piece of input data. It is used to identify and distinguish different data samples and track potential food safety risks. The signature code of the input data is calculated using a feature extraction network. If anomalies or risks exist in the data, the signature code will significantly deviate from the normal range. The method for generating a food safety signature code specifically includes the following sub-steps:

[0074] Feature extraction module: extracts local features from input data and performs weighting and normalization processing;

[0075] The input data is divided into four dimensions: production environment, product testing, supply chain logistics, and consumer feedback. Each dimension is assigned a learnable weight and a weighted fusion value is calculated to highlight key features. Specifically, the formula is used: Perform weighted fusion processing on the input data, where F represents the weighted fusion result, f i represents the eigenvalue of the i-th dimension, ω i represents the learnable weight of the i-th dimension.

[0076] Mapping module: maps the fused features to the risk interval through the activation function;

[0077] The weighted fused features are mapped to the risk interval through the activation function to enhance the sensitivity to outliers. The activation function adopts the ReLU function. Specifically, the formula: R = ReLU (F-τ) is used to represent the mapping process, where R represents the risk feature after mapping, τ represents the risk threshold, F represents the weighted fusion result, and ReLU represents the activation function.

[0078] Food safety feature code generation module: generates food safety feature codes for processed input data;

[0079] The multi-dimensional correlation of data is utilized to make the coding specific to different risks. Specifically, the formula: C = R × M is used to generate food safety feature coding, where C represents the generated food safety feature coding, R represents the mapped risk feature, M represents the sensitive adjustment matrix related to food safety, and × represents matrix multiplication.

[0080] Food Safety Risk Likelihood Module 23: Quantifying the Likelihood of Food Safety Risks in Data;

[0081] To quantify the likelihood of food safety risks in input data, a risk gradient field is introduced. This gradient field can quantify the direction and magnitude of data anomalies in the feature space. First, the risk gradient of the model output with respect to the input is calculated. The risk gradient reflects the model's sensitivity to risks near the current input. To distinguish normal fluctuations from real risks, an adaptive adjustment factor based on the distance between the input and historical safety data is introduced. This factor tends to zero when the input is close to the safe data, suppressing false positives; and tends to one when the input deviates from the safe data, exposing potential risks. Specifically, the formula is used: Calculate the possibility of food safety risk, where P represents the possibility of food safety risk, x represents the input data, and x represents the input data. safe represents historical safety data, p(y|x) represents the prediction model, and λ represents the adjustment coefficient.

[0082] Risk fusion score calculation module 24: calculates the risk fusion score based on the food safety feature code and risk probability;

[0083] The method for calculating the risk fusion score based on the food safety characteristic code and risk probability includes the following sub-steps:

[0084] Coding deviation calculation module: calculates feature coding deviation;

[0085] In order to reflect the degree of deviation between the input sample and the safe sample in the feature space, the food safety feature code C generated by the input data x is compared with the historical safe sample x. safe The generated security feature code is compared and the deviation value is calculated. Specifically, the formula is: D = || CC safe ||2 Calculate the feature coding deviation, where D represents the deviation value, C represents the food safety feature code, and C safe Represents the security feature code generated by historical security samples.

[0086] Risk fusion score calculation module: calculates the risk fusion score based on feature coding deviation and risk probability;

[0087] The risk fusion score represents a quantitative indicator of food safety risk. The larger the value, the higher the risk. Specifically, the risk fusion score is calculated using the formula: S = α·D + β·P, where S represents the risk fusion score, and α and β represent the influencing factors of feature coding deviation D and risk probability P, respectively.

[0088] Early Warning and Response Module 25: Conduct food safety risk early warning and response based on risk integration scores;

[0089] When the risk fusion score S exceeds the set safety threshold, the system automatically triggers an early warning mechanism to promptly respond to potential risks. If the high score is due to feature encoding deviations, the system immediately initiates anomaly detection processes within the production environment or supply chain to locate and isolate potentially problematic batches of products to prevent further problems. Simultaneously, the system conducts in-depth analysis of the corresponding data to identify the specific links causing the scoring anomaly and implements measures to correct the model deviations to ensure the accuracy and reliability of the scoring.

[0090] If a high score reflects the possibility of an actual risk, the system will quickly initiate laboratory re-inspections or leverage consumer feedback to verify the true source of the risk, ensuring that the risk assessment is scientifically based. At the same time, the system will automatically generate a detailed risk report, send an early warning notification to the relevant responsible departments, and initiate corresponding emergency plans based on the risk level, such as recalling problematic products, adjusting production processes, or strengthening quality control. Throughout this process, the system will continuously collect and analyze new data and dynamically optimize the risk model, thereby achieving a closed-loop risk management system and improving the system's overall prevention and control capabilities and response efficiency.

[0091] Continuous Optimization Module 26: Continuously optimize food safety data models;

[0092] By deeply mining historical data and continuously collecting real-time feedback, the system regularly updates its feature extraction network and risk quantification model, thereby continuously improving its ability to identify new risks. This process not only helps discover potential risk characteristics but also strengthens the model's adaptability and judgment in different scenarios, making risk identification more intelligent and precise, and ensuring the overall security of products and services.

[0093] The system also dynamically adjusts security thresholds and risk response strategies based on the latest industry standards, regulatory requirements, and changes in laws and policies. This mechanism allows the system to automatically adapt to changes in the external environment at different time points, ensuring the timeliness and foresight of risk prevention and control measures. While ensuring compliance, the risk management system can also continuously improve its accuracy and responsiveness, achieving comprehensive and efficient risk governance.

[0094] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent replacements, improvements, etc. made on the basis of the technical solution of the present invention should be included in the scope of protection of the present invention.

Claims

1. A food safety data operation and maintenance analysis method based on big data, characterized in that: include: S110. Collect food safety related data as input data; S120, generating a food safety characteristic code for the input data; S130, Quantify the likelihood of food safety risks in data; S140. Calculate the risk fusion score based on the food safety characteristic code and risk probability; S150. Conduct food safety risk warning and response based on risk integration scores; S160. Continuously optimize food safety data models.

2. A food safety data operation and maintenance analysis method based on big data according to claim 1, characterized in that: Collect food safety-related data as input data and deploy multi-source data collection tools, including environmental sensor networks, laboratory testing systems, logistics monitoring equipment, and public opinion analysis platforms, to automatically collect various data sources. Production environment data is transmitted to the data center in real time through IoT devices, product testing data is uploaded through laboratory information systems or rapid testing equipment, supply chain logistics data is recorded and transmitted through GPS and temperature and humidity sensors, and consumer feedback data uses natural language processing technology to extract key information from text.

3. The food safety data operation and maintenance analysis method based on big data according to claim 1, characterized in that: The method for generating food safety characteristic codes specifically includes the following sub-steps: Extract local features from input data and perform weighting and normalization; Map the fused features to the risk interval through the activation function; Generate food safety signature codes for processed input data.

4. A food safety data operation and maintenance analysis method based on big data according to claim 1, characterized in that: The calculation model outputs the risk gradient of the input. The risk gradient reflects the model's sensitivity to risks near the current input. In order to distinguish normal fluctuations from real risks, an adaptive adjustment factor based on the distance between the input and historical safety data is introduced. This factor tends to zero when the input is close to the safety data, suppressing false positives, and tends to one when the input deviates from the safety data, exposing potential risks.

5. The food safety data operation and maintenance analysis method based on big data according to claim 1, characterized in that: The method for calculating the risk fusion score based on the food safety characteristic code and risk probability includes the following sub-steps: Calculate feature encoding bias; Calculate the risk fusion score based on feature coding deviation and risk probability.

6. A food safety data operation and maintenance analysis system based on big data, characterized in that: include: Data collection module: collects food safety related data as input data; Food safety feature code generation module: generates food safety feature codes for input data; Food Safety Risk Likelihood Module: Quantifies the likelihood of food safety risks in data; Risk fusion score calculation module: calculates the risk fusion score based on food safety feature coding and risk probability; Early warning and response module: Provides food safety risk early warning and response based on risk integration scores; Continuous optimization module: Continuously optimize the food safety data model.

7. A food safety data operation and maintenance analysis system based on big data according to claim 6, characterized in that: Collect food safety-related data as input data and deploy multi-source data collection tools, including environmental sensor networks, laboratory testing systems, logistics monitoring equipment, and public opinion analysis platforms, to automatically collect various data sources. Production environment data is transmitted to the data center in real time through IoT devices, product testing data is uploaded through laboratory information systems or rapid testing equipment, supply chain logistics data is recorded and transmitted through GPS and temperature and humidity sensors, and consumer feedback data uses natural language processing technology to extract key information from text.

8. The food safety data operation and maintenance analysis system based on big data according to claim 6, characterized in that: The method for generating food safety characteristic codes specifically includes the following sub-steps: Extract local features from input data and perform weighting and normalization; Map the fused features to the risk interval through the activation function; Generate food safety signature codes for processed input data.

9. The food safety data operation and maintenance analysis system based on big data according to claim 6, characterized in that: The calculation model outputs the risk gradient of the input. The risk gradient reflects the model's sensitivity to risks near the current input. In order to distinguish normal fluctuations from real risks, an adaptive adjustment factor based on the distance between the input and historical safety data is introduced. This factor tends to zero when the input is close to the safety data, suppressing false positives, and tends to one when the input deviates from the safety data, exposing potential risks.

10. A food safety data operation and maintenance analysis system based on big data according to claim 6, characterized in that: The method for calculating the risk fusion score based on the food safety characteristic code and risk probability includes the following sub-steps: Calculate feature encoding bias; Calculate the risk fusion score based on feature coding deviation and risk probability.

Images