Logistics cargo risk assessment grading sampling inspection method, device and equipment and storage medium

Abstract

The invention relates to the technical field of information processing, in particular to a logistics cargo risk assessment grading sampling inspection method, device and equipment and a storage medium. Related report success data in a certain time period and illegal customer data recorded by prohibited goods are acquired from an upstream system; adjusting customer credit scoring information according to illegal customer data, calculating a reporting success rate based on related reporting success data, constructing a risk assessment model based on a machine learning algorithm in combination with the customer credit scoring information, the reporting success rate and historical transaction data, and performing risk scoring on each to-be-detected package by using the risk assessment model. According to the method, different risk levels can be divided according to different goods, a corresponding sampling inspection proportion and a sampling inspection process can be set for each level, pertinence and flexibility are improved, box opening sampling inspection tasks are issued according to the sampling inspection proportions and the sampling inspection processes, supervision efficiency is greatly improved, and complex market environments and rapidly changing risk conditions are effectively dealt with.

Description

Technical Field

[0001] This invention relates to the field of information processing technology, and in particular to a method, apparatus, equipment and storage medium for risk assessment and grading of logistics goods. Background Technology

[0002] Currently, with the rapid development of e-commerce and logistics, the problem of transporting prohibited items has become increasingly prominent, posing a serious threat to public safety. Existing logistics security inspection methods mainly rely on fixed frequency or random sampling, which lacks specificity and flexibility. They cannot classify different goods into different risk levels, nor can they set corresponding sampling ratios and sampling procedures for each level. This lack of specificity and flexibility makes it difficult to effectively cope with complex market environments and rapidly changing risk situations, and it is impossible to achieve automated and personalized risk assessment, thus reducing regulatory efficiency. Summary of the Invention

[0003] In order to overcome the shortcomings of the prior art, the purpose of this invention is to provide a method, device, equipment and storage medium for risk assessment and grading of logistics goods, which can classify different goods into different risk levels and set corresponding sampling ratios and sampling procedures for each level, improve targeting and flexibility, realize automated and personalized risk assessment, greatly improve regulatory efficiency, and effectively cope with complex market environments and rapidly changing risk conditions.

[0004] The first aspect of this invention provides a method for risk assessment and graded random inspection of logistics goods, comprising: acquiring relevant successful reporting data and data on customers with records of sending prohibited items from an upstream system within a certain time period; preprocessing the data on customers with violations and the relevant successful reporting data; adjusting customer credit scores based on the data on customers with violations, and calculating the reporting success rate based on the relevant successful reporting data; constructing a risk assessment model based on a machine learning algorithm combined with the customer credit scores, the reporting success rate, and historical transaction data; using the risk assessment model to score the risk of each package to be inspected, and obtaining risk level information; setting a corresponding sampling ratio and sampling process based on the risk level information; issuing unpacking and random inspection tasks to responsible customers and security personnel according to the sampling ratio and sampling process; generating a risk inspection report after the unpacking and random inspection task is marked as completed, and uploading the risk inspection report to a blockchain.

[0005] Optionally, in a first implementation of the first aspect of the present invention, the step of obtaining relevant successful reporting data and data on customers with records of sending prohibited items from the upstream system within a certain time period, and preprocessing the data on the customers with records of sending prohibited items and the relevant successful reporting data, includes: obtaining relevant successful reporting data and data on customers with records of sending prohibited items from the upstream system within a certain time period; cleaning the data on the customers with records of sending prohibited items and the relevant successful reporting data, and filling in missing values ​​for the data on the customers with records of sending prohibited items and the relevant successful reporting data; and converting the data on the customers with records of sending prohibited items and the relevant successful reporting data according to a preset target standard format.

[0006] Optionally, in a second implementation of the first aspect of the present invention, adjusting the customer credit score information based on the data of customers who violated regulations and calculating the success rate of reporting based on the relevant successful reporting data includes: performing credit scoring on all customers using a preset credit scoring model to obtain customer credit score information; adjusting the customer credit score information based on the data of customers who violated regulations; obtaining all reporting data within a certain time period from an upstream system; and calculating the success rate of reporting based on the relevant successful reporting data and all reporting data.

[0007] Optionally, in a third implementation of the first aspect of the present invention, the step of constructing a risk assessment model based on a machine learning algorithm combined with the customer credit score information, the reporting success rate, and historical transaction data, and using the risk assessment model to score the risk of each package to be inspected to obtain risk level information, includes: collecting historical transaction data; extracting key features of the customer credit score information, the reporting success rate, and the historical transaction data; constructing a risk assessment model based on a machine learning algorithm combined with the key features; and using the risk assessment model to score the risk of each package to be inspected to obtain risk level information, wherein the risk level information includes high risk, medium risk, and low risk.

[0008] Optionally, in a fourth implementation of the first aspect of the present invention, the step of setting a corresponding sampling ratio and sampling process based on the risk level information, and issuing unpacking sampling tasks to responsible customers and security personnel according to the sampling ratio and sampling process, includes: calling a preset sampling plan; matching a sampling ratio and sampling process with a mapping relationship from the sampling plan based on the risk level information; and issuing unpacking sampling tasks to responsible customers and security personnel according to the sampling ratio and sampling process.

[0009] Optionally, in a fifth implementation of the first aspect of the present invention, the step of generating a risk inspection report and uploading the risk inspection report to the blockchain after the unpacking and inspection task is marked as completed includes: obtaining inspection information after the unpacking and inspection task is marked as completed; generating a risk inspection report based on the risk level information and the inspection information; encrypting the risk inspection report to obtain an encrypted risk inspection report; and uploading the encrypted risk inspection report to the blockchain.

[0010] Optionally, in a sixth implementation of the first aspect of the present invention, after generating a risk inspection report and uploading the risk inspection report to the blockchain after the unpacking and sampling inspection task is marked, the method further includes: analyzing the accuracy of the classification based on the risk inspection report to obtain analysis results; generating model adjustment parameters based on the analysis results; and optimizing the risk assessment model based on the model adjustment parameters.

[0011] A second aspect of the present invention provides a risk assessment and grading sampling inspection device for logistics goods, comprising: an acquisition and processing module, used to acquire relevant successful reporting data and data of customers with records of sending prohibited items from an upstream system within a certain time period, and preprocess the data of customers with violations and the relevant successful reporting data; an adjustment and calculation module, used to adjust customer credit score information according to the data of customers with violations, and calculate the reporting success rate based on the relevant successful reporting data; a scoring construction module, used to construct a risk assessment model based on a machine learning algorithm combined with the customer credit score information, the reporting success rate and historical transaction data, and use the risk assessment model to score the risk of each package to be inspected to obtain risk level information; a setting and distribution module, used to set the corresponding sampling ratio and sampling process according to the risk level information, and distribute the unpacking sampling inspection task to the responsible customer and security personnel according to the sampling ratio and sampling process; and a generation and uploading module, used to generate a risk sampling inspection report after the unpacking sampling inspection task is marked, and upload the risk sampling inspection report to the blockchain.

[0012] Optionally, in a first implementation of the second aspect of the present invention, the acquisition and processing module includes: a first acquisition unit, configured to acquire relevant successful reporting data and violation customer data with records of sending prohibited items from an upstream system within a certain time period; a cleaning and filling unit, configured to clean the violation customer data and the relevant successful reporting data, and fill in missing values ​​in the violation customer data and the relevant successful reporting data; and a conversion unit, configured to convert the violation customer data and the relevant successful reporting data into a format based on a preset target standard format.

[0013] Optionally, in a second implementation of the second aspect of the present invention, the adjustment calculation module includes: a first scoring unit, used to perform credit scoring on all customers using a preset credit scoring model to obtain customer credit scoring information; an adjustment unit, used to adjust the customer credit scoring information according to the data of the violating customers; a second acquisition unit, used to acquire all reporting data within a certain time period from the upstream system; and a calculation unit, used to calculate the reporting success rate based on the relevant successful reporting data and all reporting data.

[0014] Optionally, in a third implementation of the second aspect of the present invention, the scoring module includes: a collection unit for collecting historical transaction data; an extraction unit for extracting the customer credit score information, the reporting success rate, and key features of the historical transaction data; a construction unit for constructing a risk assessment model based on a machine learning algorithm combined with the key features; and a second scoring unit for using the risk assessment model to score the risk of each package to be inspected, thereby obtaining risk level information, wherein the risk level information includes high risk, medium risk, and low risk.

[0015] Optionally, in a fourth implementation of the second aspect of the present invention, the setting and issuing module includes: a calling unit for calling a preset sampling plan; a matching unit for matching a sampling ratio and sampling process with a mapping relationship from the sampling plan according to the risk level information; and an issuing unit for issuing unpacking sampling tasks to responsible customers and security personnel according to the sampling ratio and sampling process.

[0016] Optionally, in a fifth implementation of the second aspect of the present invention, the generation and uploading module includes: a third acquisition unit, configured to acquire sampling information after the unpacking and sampling task is marked; a generation unit, configured to generate a risk sampling report based on the risk level information and the sampling information; an encryption unit, configured to encrypt the risk sampling report to obtain an encrypted risk sampling report; and an uploading unit, configured to upload the encrypted risk sampling report to the blockchain.

[0017] Optionally, in a sixth implementation of the second aspect of the present invention, the method further includes: an analysis module, used to analyze the accuracy of the classification based on the risk sampling report and obtain analysis results; a generation module, used to generate model adjustment parameters according to the analysis results; and an optimization module, used to optimize the risk assessment model according to the model adjustment parameters.

[0018] A third aspect of the present invention provides a risk assessment and grading sampling inspection device for logistics goods, the device comprising: a memory and at least one processor, the memory storing instructions; at least one processor calling the instructions in the memory to cause the risk assessment and grading sampling inspection device for logistics goods to perform each step of the risk assessment and grading sampling inspection method for logistics goods as described above.

[0019] A fourth aspect of the present invention provides a computer-readable storage medium storing instructions that, when executed by a processor, implement the steps of the logistics cargo risk assessment and grading sampling method described in any of the preceding claims.

[0020] In the technical solution of this invention, relevant successful reporting data and data on customers with records of sending prohibited items are obtained from the upstream system within a certain period of time. Customer credit scores are adjusted based on the data on customers with violations. The success rate of reporting is calculated based on relevant successful reporting data. A risk assessment model is constructed based on machine learning algorithms, combined with customer credit scores, reporting success rates, and historical transaction data. The risk assessment model is used to score the risk of each package to be inspected. Different goods can be divided into different risk levels, and corresponding sampling ratios and sampling procedures can be set for each level, improving targeting and flexibility, and realizing automated and personalized risk assessment. The task of opening and inspecting packages is issued to responsible customers and security personnel according to the sampling ratio and sampling procedure, which greatly improves the efficiency of supervision and effectively responds to complex market environments and rapidly changing risk situations. Attached Figure Description

[0021] Figure 1 A first flowchart of the risk assessment and grading sampling inspection method for logistics goods provided in this embodiment of the invention;

[0022] Figure 2 This is a second flowchart of the risk assessment and grading sampling method for logistics goods provided in an embodiment of the present invention;

[0023] Figure 3 This is a third flowchart of the risk assessment and grading sampling method for logistics goods provided in this embodiment of the invention;

[0024] Figure 4 The fourth flowchart of the risk assessment and grading sampling method for logistics goods provided in this embodiment of the invention;

[0025] Figure 5 A schematic diagram of a logistics cargo risk assessment and grading sampling inspection device provided in an embodiment of the present invention;

[0026] Figure 6This is another structural schematic diagram of the logistics cargo risk assessment and grading sampling inspection device provided in an embodiment of the present invention;

[0027] Figure 7 This is a schematic diagram of the structure of the logistics cargo risk assessment and grading sampling inspection equipment provided in an embodiment of the present invention. Detailed Implementation

[0028] This invention provides a method, apparatus, equipment, and storage medium for risk assessment and tiered sampling of logistics goods. It can classify different goods into different risk levels and set corresponding sampling ratios and sampling procedures for each level, thereby improving targeting and flexibility, realizing automated and personalized risk assessment, greatly improving regulatory efficiency, and effectively responding to complex market environments and rapidly changing risk conditions.

[0029] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” or “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0030] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1 One embodiment of the risk assessment and grading sampling method for logistics goods in this invention includes:

[0031] 101. Obtain relevant successful reporting data and data on customers with records of sending prohibited items from the upstream system within a certain time period, and preprocess the data on customers with violations and relevant successful reporting data;

[0032] In this embodiment, the time period for data acquisition is clearly defined. Appropriate query tools (SQL, API requests, etc.) are used to obtain relevant successful reporting data and data of violating customers with records of sending prohibited items from the upstream system within a certain time period. The system checks whether there are duplicate items in the data and removes duplicate data. For numerical data, the mean, median or other statistical values ​​can be used to fill the data. For categorical data (such as customer behavior type, reporting category, etc.), pattern values ​​can be used to fill the data.

[0033] 102. Adjust customer credit scores based on data of customers who violate regulations, and calculate the success rate of reports based on relevant successful reporting data;

[0034] In this embodiment, a customer's initial credit score can be calculated based on indicators such as their historical behavior, financial status, and customer type. For customers who violate regulations, especially those involved in repeated or serious violations, the credit score is adjusted using a weighted coefficient. For example, if a customer has a successful report record and violates regulations multiple times, their credit score will decrease. The adjusted credit score will reflect the customer's risk level. Customers with low scores may require additional review or risk management. The success rate of a report usually refers to the proportion of successful reports out of the total number of reports within a certain period of time.

[0035] 103. Based on machine learning algorithms, a risk assessment model is constructed by combining customer credit scores, reporting success rates, and historical transaction data. The risk assessment model is used to score the risk of each package to be inspected, and the risk level information is obtained.

[0036] In this embodiment, data from different sources, including customer credit scores, reporting success rates, and historical transaction data, are integrated as inputs to train a machine learning model. A suitable machine learning algorithm is selected to train the risk assessment model. Based on the integrated data, an accurate risk assessment model is trained using the machine learning algorithm. The trained risk assessment model is then used to score each package to be inspected to obtain risk level information.

[0037] 104. Set the corresponding sampling ratio and sampling process according to the risk level information, and issue the unpacking and sampling task to the responsible customers and security personnel in accordance with the sampling ratio and sampling process.

[0038] In this embodiment, by classifying risks into levels, the proportion of packages requiring random inspection for each risk level is determined, ensuring that high-risk packages receive priority attention without affecting the processing efficiency of low-risk packages. Based on the risk level information of the packages, the system automatically generates random inspection tasks and assigns them to responsible customers and security personnel. The task content should include the specific packages to be inspected, the time requirements for inspection, and the inspection objectives. Through a standardized and systematic inspection process, the inspection work is ensured to be carried out efficiently and orderly, while reducing human error and omissions.

[0039] 105. After the unpacking and sampling inspection task is marked as completed, generate a risk sampling inspection report and upload the risk sampling inspection report to the blockchain;

[0040] In this embodiment, the completion status of each sampling inspection task should be confirmed by the executor (security personnel) or the responsible customer. After confirmation, the task status is marked as "completed" through the system interface, and relevant sampling inspection results, time, executor and other information are attached. After the unpacking sampling inspection task is marked as completed, a detailed risk sampling inspection report is generated, recording the execution status, results and potential risks found for each sampling inspection task, and the risk sampling inspection report is uploaded to the blockchain.

[0041] In this embodiment of the invention, relevant successful reporting data and data on customers with records of sending prohibited items are obtained from the upstream system within a certain period. Customer credit scores are adjusted based on the data of customers with violations. The success rate of reporting is calculated based on relevant successful reporting data. A risk assessment model is constructed based on machine learning algorithms, combined with customer credit scores, reporting success rates, and historical transaction data. The risk assessment model is used to score the risk of each package to be inspected. Different goods can be divided into different risk levels, and corresponding sampling ratios and sampling procedures can be set for each level, improving targeting and flexibility, and realizing automated and personalized risk assessment. The task of opening and sampling inspection is issued to responsible customers and security personnel according to the sampling ratio and sampling procedure, which greatly improves the efficiency of supervision and effectively responds to complex market environments and rapidly changing risk situations.

[0042] Please see Figure 2 The second embodiment of the risk assessment and grading sampling method for logistics goods in this invention includes:

[0043] 201. Obtain relevant successful reporting data and data on customers with records of sending prohibited items from the upstream system within a certain period of time;

[0044] In this embodiment, a time range (such as the past 30 days, the past 7 days, etc.) is defined as the standard to filter and obtain relevant successful reporting data and data of customers who have a record of sending prohibited items.

[0045] 202. Perform data cleaning on the data of customers who violated regulations and the data of successful reports, and fill in the missing values ​​in the data of customers who violated regulations and the data of successful reports.

[0046] In this embodiment, it checks whether a customer appears multiple times in the data, checks whether there are duplicate report records, removes duplicate data of violating customer data and related successful report data, identifies the specific location and type of missing values, selects different filling strategies for different fields, and fills missing values ​​in violating customer data and related successful report data according to different filling strategies.

[0047] 203. Based on the preset target standard format, convert the format of data on customers who violated regulations and related successful reports;

[0048] In this embodiment, the requirements of the target standard format are clearly defined. Based on the preset target standard format, the format conversion of the data of customers who violated regulations and the data of successful reports is performed. The consistency of field names is specified to avoid inconsistencies in field naming between different systems. The data type of each field is clearly defined, such as string, integer, date and time, to ensure that the field order in the data is consistent with the target format.

[0049] 204. Use the preset credit scoring model to score all customers and obtain customer credit score information;

[0050] In this embodiment, the credit scoring model is built based on a set of known customer characteristics and behavioral data through statistical methods, machine learning algorithms, or expert experience. Sufficiently comprehensive customer data is collected and input into the credit scoring model. The credit scoring model calculates the credit score of each customer based on the input customer data and a preset calculation formula or algorithm (such as the regression equation in logistic regression or the prediction function of a machine learning model) and outputs it to obtain customer credit score information.

[0051] 205. Adjust customer credit scores based on data on customers who have violated regulations;

[0052] In this embodiment, the customer credit score information is adjusted based on the data of customers who violated regulations. For example, the credit score information of customers who violated regulations will be lowered.

[0053] 206. Obtain all reported data within a certain time period from the upstream system;

[0054] In this embodiment, the time period for the reported data is clearly defined, such as monthly, quarterly, or annually. The upstream system provides an API interface that allows the reported data to be obtained through programmatic calls, thereby retrieving all reported data within a certain time period from the upstream system.

[0055] 207. Calculate the reporting success rate based on relevant successful reporting data and all reporting data;

[0056] In this embodiment, the reporting success rate is a quantification of the relationship between the effective processing results of all reported data and the total number of reports. A successful report means that the report is confirmed as valid after evaluation, while an unsuccessful report means that the report is not confirmed as valid, or the content of the report is untrue or lacks supporting information, or the report does not match the actual situation and fails to have an actual impact. First, all reported data within a certain period of time is obtained from the upstream system, and all successful reported data is filtered out. The reporting success rate is calculated based on the relevant successful reported data and all reported data.

[0057] In this embodiment of the invention, by obtaining data from different sources (such as successful report data and data on customers who violated regulations) from upstream systems, the comprehensiveness and diversity of the data are ensured, which helps to build a more comprehensive and accurate model. Steps such as data cleaning, missing value imputation, and format conversion ensure the quality and consistency of the data, reduce erroneous analysis and decision-making caused by data problems, and ensure the accuracy and effectiveness of subsequent processing. By scoring customers through a credit scoring model and adjusting it in conjunction with violation records, it helps to accurately reflect the risk status of customers and improve the accuracy of risk assessment. The report data and data on customers who violated regulations obtained from upstream systems can achieve real-time dynamic monitoring, ensuring that problems are detected in a timely manner and corresponding adjustments are made. Calculating the success rate of reports and making corresponding score adjustments helps to optimize customer management strategies, improve the efficiency and quality of decision-making, and provide strong support for risk management.

[0058] Please see Figure 3 The third embodiment of the risk assessment and grading sampling method for logistics goods in this invention includes:

[0059] 301. Collect historical transaction data;

[0060] In this embodiment, historical transaction data is automatically obtained through the API interface provided by the trading platform, and duplicate transaction records in the historical transaction data are removed.

[0061] 302. Extract key features from customer credit scoring information, complaint success rate, and historical transaction data;

[0062] In this embodiment, key features of customer credit score information, key features of reporting success rate, and key features of historical transaction data are extracted, and the key features of credit score information, reporting success rate, and historical transaction data are merged.

[0063] 303. Construct a risk assessment model based on machine learning algorithms and key features;

[0064] In this embodiment, the collected data is ensured to be sufficiently representative of different risk factors. Key features related to risk are selected. Based on the characteristics of the data and the model's objectives, a suitable machine learning algorithm is chosen. The selected machine learning algorithm and key features are used to train the risk assessment model. The model's hyperparameters are adjusted using methods such as grid search and random search to optimize model performance. Ensemble learning methods (such as random forest and XGBoost) can be used to improve the model's stability and accuracy. The optimized model is then deployed to a real-world environment.

[0065] 304. Use a risk assessment model to score the risk of each package to be inspected and obtain risk level information, which includes high risk, medium risk and low risk.

[0066] In this embodiment, for each package to be inspected, all relevant feature data are collected, and the numerical features are normalized or standardized to ensure that the scale of different features is consistent. The category features (such as package type, region, etc.) are converted into a numerical form suitable for model input. The features extracted above are used as input, and the previously constructed risk assessment model is used to score them. The model will predict each package based on the input features and output a risk level information, which is one of high risk, medium risk, and low risk.

[0067] 305. Invoke the preset sampling plan;

[0068] In this embodiment, different sampling strategies can be pre-designed according to the risk level of the package, thereby obtaining a preset sampling plan. When the risk assessment model outputs the risk level of the package, the preset sampling plan can be called through an automated system.

[0069] 306. Match the sampling ratio and sampling process with a mapping relationship from the sampling plan based on the risk level information;

[0070] In this embodiment, after receiving the risk level information, the system automatically determines the sampling ratio and corresponding specific process for the package by looking up the mapping table in the sampling plan. The sampling plan predefines the sampling ratio corresponding to different risk levels, and each risk level corresponds to a specific sampling process. The process also defines the sampling timing, operation steps, required equipment and personnel configuration, etc. For example, if the risk level is high, the ratio and process corresponding to high risk are directly called.

[0071] 307. Issue unpacking and sampling inspection tasks to responsible customers and security personnel in accordance with the sampling ratio and sampling process;

[0072] In this embodiment, according to the matching sampling inspection process, the system issues specific execution instructions to the on-site operators, including detailed operation guidelines such as inspection steps, equipment used, and recording methods. The sampling inspection results and process execution status are uploaded to the system in real time for subsequent analysis and improvement.

[0073] In this embodiment of the invention, by collecting historical transaction data and extracting key features, and combining them with machine learning algorithms for risk assessment, valuable information can be extracted from a large amount of historical data, enabling accurate risk prediction and assessment. Based on the automatic matching of risk scores and sampling plans, the intelligence level of the sampling process is improved, manual intervention is reduced, and work efficiency and accuracy are enhanced. By classifying risks into high-risk, medium-risk, and low-risk categories, high-risk packages can be prioritized, resources can be allocated rationally, and potential risks can be effectively controlled. By selecting the corresponding sampling ratio and process according to the risk level, the precision and compliance of the sampling work are ensured, management efficiency and risk control effectiveness are improved, and the final task assignment step ensures clear responsibility allocation. Through the collaboration between security personnel and customers, the smooth execution of the sampling process is guaranteed, enhancing the collaborative efficiency of the entire process.

[0074] Please see Figure 4 The fourth embodiment of the risk assessment and grading sampling method for logistics goods in this invention includes:

[0075] 401. Obtain sampling inspection information after the unpacking and sampling inspection task is marked as completed;

[0076] In this embodiment, during the unpacking and random inspection task, the responsible customer and security personnel need to strictly follow the procedure to conduct random inspections and record the inspection results of each package. After the task is completed, the responsible customer and security personnel mark the random inspection task as "completed" through the system interface. This operation includes filling in the inspection results (qualified, unqualified, or pending verification, etc.) and uploading relevant records (photos, reports, signatures, etc.). After the system confirms that the task is marked as "completed", the system will enter the automatic information collection and summary stage, and the system will automatically extract all information related to the completion of the task.

[0077] 402. Generate a risk sampling inspection report based on risk level information and sampling inspection results;

[0078] In this embodiment, the sampling inspection information includes the specific sampling inspection results of each package, such as qualified, unqualified, or pending verification. The risk level information and the sampling inspection information are combined, that is, the sampling inspection results and risk level information of each package are summarized to obtain a risk sampling inspection report.

[0079] 403. Encrypt the risk sampling inspection report to obtain an encrypted risk sampling inspection report;

[0080] In this embodiment, a specific encryption algorithm is selected to encrypt the risk sampling report, such as a symmetric encryption algorithm, an asymmetric encryption algorithm, or a hash algorithm, and the encrypted risk sampling report is obtained after encryption.

[0081] 404. Upload the encrypted risk inspection report to the blockchain;

[0082] In this embodiment, a suitable blockchain platform, such as Ethereum, Hyperledger Fabric, or EOS, is selected. The encrypted report is hashed, and a smart contract is created to handle the upload operation. The smart contract interface is called to transmit information such as the hash value and report ID to the blockchain. Once the transaction is successful, the blockchain network records the hash value and timestamp of the report and generates a block. Nodes in the blockchain network verify the transaction. After confirming that it is correct, they package the transaction into a block and append it to the blockchain. After successful upload, all nodes in the blockchain network have the hash information of the report.

[0083] 405. Analyze the accuracy of the classification based on the risk sampling report and obtain the analysis results;

[0084] In this embodiment, an appropriate analysis model is selected to evaluate the hierarchical accuracy based on the nature of the data and the complexity of the task, such as support vector machine, random forest and neural network. Based on the evaluation results of the model, detailed analysis results are generated, including detailed explanations of evaluation indicators such as accuracy, precision and recall.

[0085] 406. Generate a model and adjust parameters based on the analysis results;

[0086] In this embodiment, the analysis results are interpreted, and based on the interpretation results, the model parameters are adjusted or different tuning strategies are selected. New features are created based on the tuning strategies to obtain the model adjustment parameters.

[0087] 407. Optimize the risk assessment model by adjusting parameters according to the model;

[0088] In this embodiment, the risk assessment model is retrained by adjusting the model parameters, and key indicators such as loss and accuracy are recorded during the training process to ensure consistent performance on the training and validation sets. K-fold cross-validation is used to verify the robustness of the model and reduce the risk of overfitting to a specific training set. The model is re-evaluated using indicators such as confusion matrix, AUC, precision, and recall to ensure that the adjusted model can effectively identify high-risk events and perform well in multiple evaluation indicators. The performance of the model before and after adjustment is compared to check whether there is a significant improvement in various indicators (such as precision, recall, AUC, etc.). The performance of the optimized model is summarized and an analysis report is generated. The report includes the parameter adjustment process, the performance of the adjusted model, and the improvement effect of risk assessment.

[0089] In this embodiment of the invention, by uploading encrypted risk sampling reports to the blockchain, the security, immutability, and transparency of the data are ensured, increasing the credibility of the system and the reliability of the data. Through precise analysis and adjustment of the risk assessment model, a closed-loop system based on data feedback is formed. This feedback mechanism continuously optimizes the model, improving its accuracy and effectiveness. By analyzing the risk sampling reports, the accuracy of risk level assessment can be precisely determined, and by optimizing model parameters, the accuracy of the decision support system can be further improved. From report generation to uploading to model optimization, the entire process is highly automated, which not only improves work efficiency but also reduces the possibility of human intervention, ensuring the smoothness and consistency of the process. With the help of blockchain technology, all risk assessment and adjustment processes are traceable, ensuring compliance and auditing convenience.

[0090] The method for risk assessment and tiered sampling inspection of logistics goods in this embodiment of the invention has been described above. The following describes the device for risk assessment and tiered sampling inspection of logistics goods in this embodiment of the invention. Please refer to [link / reference]. Figure 5 One embodiment of the logistics cargo risk assessment and grading sampling inspection device of the present invention includes:

[0091] The acquisition and processing module 501 is used to acquire relevant successful reporting data and data of customers with records of sending prohibited items from the upstream system within a certain time period, and to preprocess the data of customers with violations and relevant successful reporting data.

[0092] The calculation module 502 is adjusted to adjust the customer credit score information based on the data of customers who violated regulations, and to calculate the success rate of reports based on relevant successful report data.

[0093] A scoring module 503 is constructed to build a risk assessment model based on machine learning algorithms, customer credit scores, reporting success rates, and historical transaction data. The risk assessment model is used to score the risk of each package to be inspected and obtain risk level information.

[0094] The setting and distribution module 504 is used to set the corresponding sampling ratio and sampling process according to the risk level information, and to issue the unpacking and sampling task to the responsible customers and security personnel according to the sampling ratio and sampling process.

[0095] The upload module 505 is used to generate a risk inspection report after the unpacking and sampling inspection task is marked as completed, and then upload the risk inspection report to the blockchain.

[0096] In this embodiment, relevant successful reporting data and data on customers with records of sending prohibited items are obtained from the upstream system within a certain period. Customer credit scores are adjusted based on the data of customers with violations. The success rate of reporting is calculated based on relevant successful reporting data. A risk assessment model is constructed based on machine learning algorithms, combined with customer credit scores, reporting success rates, and historical transaction data. The risk assessment model is used to score the risk of each package to be inspected. Different goods can be divided into different risk levels, and corresponding sampling ratios and sampling procedures can be set for each level, improving targeting and flexibility, and realizing automated and personalized risk assessment. Opening and sampling tasks are issued to responsible customers and security personnel according to the sampling ratios and sampling procedures, which greatly improves regulatory efficiency and effectively responds to complex market environments and rapidly changing risk situations.

[0097] Please see Figure 6 Another embodiment of the logistics cargo risk assessment and grading sampling inspection device of the present invention includes:

[0098] The acquisition and processing module 501 is used to acquire relevant successful reporting data and data of customers with records of sending prohibited items from the upstream system within a certain time period, and to preprocess the data of customers with violations and relevant successful reporting data.

[0099] The calculation module 502 is adjusted to adjust the customer credit score information based on the data of customers who violated regulations, and to calculate the success rate of reports based on relevant successful report data.

[0100] A scoring module 503 is constructed to build a risk assessment model based on machine learning algorithms, customer credit scores, reporting success rates, and historical transaction data. The risk assessment model is used to score the risk of each package to be inspected and obtain risk level information.

[0101] The setting and distribution module 504 is used to set the corresponding sampling ratio and sampling process according to the risk level information, and to issue the unpacking and sampling task to the responsible customers and security personnel according to the sampling ratio and sampling process.

[0102] The upload module 505 is used to generate a risk inspection report after the unpacking and sampling inspection task is marked as completed, and upload the risk inspection report to the blockchain.

[0103] In this embodiment, the acquisition and processing module 501 includes: a first acquisition unit 5011, used to acquire relevant successful reporting data and data of customers with records of sending prohibited items from the upstream system within a certain time period; a cleaning and filling unit 5012, used to clean the data of customers with violations and relevant successful reporting data, and fill in missing values ​​in the data of customers with violations and relevant successful reporting data; and a conversion unit 5013, used to convert the data of customers with violations and relevant successful reporting data into a format based on a preset target standard format.

[0104] In this embodiment, the adjustment calculation module 502 includes: a first scoring unit 5021, used to score all customers' credit scores using a preset credit scoring model to obtain customer credit score information; an adjustment unit 5022, used to adjust the customer credit score information based on the data of customers who violated regulations; a second acquisition unit 5023, used to acquire all report data within a certain time period from the upstream system; and a calculation unit 5024, used to calculate the report success rate based on relevant successful report data and all report data.

[0105] In this embodiment, the scoring module 503 includes: a collection unit 5031 for collecting historical transaction data; an extraction unit 5032 for extracting customer credit score information, reporting success rate, and key features of historical transaction data; a construction unit 5033 for constructing a risk assessment model based on machine learning algorithms and key features; and a second scoring unit 5034 for using the risk assessment model to score the risk of each package to be inspected, thereby obtaining risk level information, which includes high risk, medium risk, and low risk.

[0106] In this embodiment, the setting and issuing module 504 includes: a calling unit 5041, used to call a preset sampling plan; a matching unit 5042, used to match the sampling ratio and sampling process with a mapping relationship from the sampling plan according to the risk level information; and an issuing unit 5043, used to issue the unpacking sampling task to the responsible customer and security personnel according to the sampling ratio and sampling process.

[0107] In this embodiment, the generation and upload module 505 includes: a third acquisition unit 5051, used to acquire sampling information after the unpacking and sampling inspection task is marked; a generation unit 5052, used to generate a risk sampling report based on risk level information and sampling information; an encryption unit 5053, used to encrypt the risk sampling report to obtain an encrypted risk sampling report; and an upload unit 5054, used to upload the encrypted risk sampling report to the blockchain.

[0108] In this embodiment, the system further includes: an analysis module 506, used to analyze the accuracy of the classification based on the risk sampling report and obtain analysis results; a generation module 507, used to generate model adjustment parameters based on the analysis results; and an optimization module 508, used to optimize the risk assessment model based on the model adjustment parameters.

[0109] above Figure 5 and Figure 6 The logistics cargo risk assessment and grading sampling inspection device in this embodiment of the invention is described in detail from the perspective of modular functional entities. The logistics cargo risk assessment and grading sampling inspection device in this embodiment of the invention is described in detail from the perspective of hardware processing.

[0110] Figure 7 This is a schematic diagram of a logistics cargo risk assessment and grading sampling inspection device 600 provided in an embodiment of the present invention. The logistics cargo risk assessment and grading sampling inspection device 600 can vary significantly due to different configurations or performance. It may include one or more central processing units (CPUs) 610 (e.g., one or more processors) and a memory 620, and one or more storage media 630 (e.g., one or more mass storage devices) storing application programs 633 or data 632. The memory 620 and storage media 630 can be temporary or persistent storage. The program stored in the storage media 630 may include one or more modules (not shown in the diagram), each module may include a series of instruction operations on the logistics cargo risk assessment and grading sampling inspection device 600. Furthermore, the processor 610 may be configured to communicate with the storage media 630 and execute the series of instruction operations in the storage media 630 on the logistics cargo risk assessment and grading sampling inspection device 600 to implement the steps of the logistics cargo risk assessment and grading sampling inspection method provided in the above-described method embodiments.

[0111] The logistics cargo risk assessment and grading sampling inspection equipment 600 may also include one or more power supplies 640, one or more wired or wireless network interfaces 650, one or more input / output interfaces 660, and / or one or more operating systems 631, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will understand that... Figure 7 The illustrated structure of the logistics cargo risk assessment and grading sampling inspection equipment does not constitute a limitation on the logistics cargo risk assessment and grading sampling inspection equipment. It may include more or fewer components than illustrated, or combine certain components, or have different component arrangements.

[0112] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when executed on a computer, cause the computer to perform the steps of the logistics cargo risk assessment and grading sampling inspection method.

[0113] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system, device, or unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0114] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0115] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for risk assessment and tiered sampling inspection of logistics goods, characterized in that, include: The system obtains relevant successful reporting data and data on customers with records of sending prohibited items from the upstream system within a certain time period, and preprocesses the data on the violating customers and the relevant successful reporting data. The customer credit score information is adjusted based on the aforementioned data on customers who violated regulations, and the success rate of reporting is calculated based on the relevant successful reporting data. A risk assessment model is constructed based on machine learning algorithms, combined with the customer credit score information, the reporting success rate and historical transaction data. The risk assessment model is used to score the risk of each package to be inspected to obtain risk level information. Based on the risk level information, set the corresponding sampling ratio and sampling process, and issue the unpacking and sampling task to the responsible customers and security personnel in accordance with the sampling ratio and sampling process. After the unpacking and sampling inspection task is marked as completed, a risk sampling inspection report is generated and uploaded to the blockchain.

2. The method for risk assessment and tiered sampling inspection of logistics goods according to claim 1, characterized in that, The process involves obtaining relevant successful reporting data and data on customers with records of sending prohibited items from the upstream system within a certain time period, and preprocessing the data on the violating customers and the relevant successful reporting data, including: Obtain relevant successful reporting data and violation customer data with records of sending prohibited items from the upstream system within a certain period of time; The data on the violating customers and the related successful reporting data are cleaned, and missing values ​​are filled in for the data on the violating customers and the related successful reporting data. Based on a preset target standard format, the data of customers who violated regulations and the data of successful related reports are converted into a new format.

3. The method for risk assessment and tiered sampling inspection of logistics goods according to claim 1, characterized in that, The step of adjusting customer credit scores based on the data of customers who violated regulations, and calculating the success rate of reports based on the relevant successful reporting data, includes: The system uses a pre-defined credit scoring model to score all customers and obtain their credit score information. Adjust the customer credit score information based on the aforementioned data on customers who violated regulations; Obtain all reported data within a certain time period from the upstream system; The success rate of a report is calculated based on the relevant successful report data and all the reported data.

4. The method for risk assessment and tiered sampling inspection of logistics goods according to claim 1, characterized in that, The risk assessment model, constructed based on machine learning algorithms and incorporating customer credit scores, reporting success rates, and historical transaction data, is used to score the risk of each package to be inspected, yielding risk level information, including: Collect historical transaction data; Extract key features from the customer credit score information, the report success rate, and the historical transaction data; A risk assessment model is constructed based on machine learning algorithms and the aforementioned key features; The risk assessment model is used to score the risk of each package to be inspected, and the risk level information includes high risk, medium risk and low risk.

5. The method for risk assessment and tiered sampling inspection of logistics goods according to claim 1, characterized in that, The process of setting corresponding sampling ratios and procedures based on the risk level information, and issuing unpacking and sampling inspection tasks to responsible customers and security personnel according to the sampling ratios and procedures, includes: Invoke the preset sampling plan; Based on the risk level information, a sampling ratio and sampling process with a mapping relationship are matched from the sampling plan; In accordance with the aforementioned sampling ratio and sampling process, the task of unpacking and sampling inspection was issued to the responsible customers and security personnel.

6. The method for risk assessment and tiered sampling inspection of logistics goods according to claim 1, characterized in that, The process of generating a risk inspection report after the unpacking and sampling inspection task is marked as completed, and uploading the risk inspection report to the blockchain, includes: After the unpacking and sampling inspection task is marked as completed, obtain the sampling inspection information; A risk sampling report is generated based on the risk level information and the sampling information. The risk sampling report is encrypted to obtain an encrypted risk sampling report. The encrypted risk sampling report is uploaded to the blockchain.

7. The method for risk assessment and tiered sampling inspection of logistics goods according to claim 1, characterized in that, After the unpacking and sampling inspection task is marked as completed, a risk sampling inspection report is generated, and the risk sampling inspection report is uploaded to the blockchain, the process further includes: The accuracy of the classification was analyzed based on the aforementioned risk sampling report, and the analysis results were obtained. Based on the analysis results, generate a model and adjust the parameters; The risk assessment model is optimized by adjusting the parameters according to the model.

8. A risk assessment and grading sampling inspection device for logistics goods, characterized in that, include: The acquisition and processing module is used to acquire relevant successful reporting data and data of customers with records of sending prohibited items from the upstream system within a certain time period, and to preprocess the data of the violating customers and the relevant successful reporting data. The adjustment calculation module is used to adjust the customer credit score information based on the data of customers who violated regulations, and to calculate the success rate of reports based on the relevant successful report data; A scoring module is constructed to build a risk assessment model based on machine learning algorithms, combined with the customer credit score information, the reporting success rate and historical transaction data. The risk assessment model is used to score the risk of each package to be inspected to obtain risk level information. The setting and distribution module is used to set the corresponding sampling ratio and sampling process according to the risk level information, and to issue the unpacking and sampling task to the responsible customers and security personnel according to the sampling ratio and sampling process. The generation and upload module is used to generate a risk inspection report after the unpacking and sampling inspection task is marked, and upload the risk inspection report to the blockchain.

9. A risk assessment and grading sampling inspection device for logistics goods, characterized in that, The logistics cargo risk assessment and grading sampling inspection equipment includes: a memory and at least one processor, wherein the memory stores instructions; At least one of the processors invokes the instructions in the memory to cause the logistics cargo risk assessment and grading sampling inspection device to perform the steps of the logistics cargo risk assessment and grading sampling inspection method as described in any one of claims 1-7.

10. A computer-readable storage medium storing instructions thereon, characterized in that, When the instructions are executed by the processor, they implement each step of the logistics cargo risk assessment and grading sampling method as described in any one of claims 1-7.

Images