Customs declaration document identification-based customs risk detection method and system

Abstract

The application provides a customs clearance risk detection method and system based on customs declaration document recognition, and relates to the technical field of customs declaration document risk management. The method comprises the following steps: obtaining a scanned copy of a basic customs declaration document, and determining whether additional customs declaration documents are required; if additional customs declaration documents are required, generating upload prompt information; determining a high customs declaration risk trigger result according to the recognition result of the scanned copy of the basic customs declaration document and the upload result of the scanned copy of the additional customs declaration document; determining a medium customs declaration risk trigger result according to the recognition results of the scanned copy of the basic customs declaration document and the scanned copy of the additional customs declaration document; determining a low customs declaration risk trigger result according to the recognition result of the scanned copy of the basic customs declaration document; and determining a customs clearance risk detection result according to the high customs declaration risk trigger result, the medium customs declaration risk trigger result and the low customs declaration risk trigger result. According to the application, accurate prompts can be given based on the risk level of the customs declaration documents, so as to improve the probability of smooth customs clearance.

Description

Technical Field

[0001] This invention relates to the field of customs declaration risk management technology, and in particular to a method and system for detecting customs clearance risks based on customs declaration document identification. Background Technology

[0002] In related technologies, CN119228246A relates to the field of data processing technology. More specifically, this solution relates to an intelligent analysis method for customs declaration data, including: obtaining the export quantity and price of all customs declarations for any item from any customer; calculating the data similarity between each customs declaration and the remaining customs declarations; constructing a first coordinate system; determining representative quantity and representative price; calculating the deviation degree of each customs declaration; constructing a second coordinate system; calculating the weighted Euclidean distance between feature points of each customs declaration and any other customs declaration; and obtaining the anomaly score of each customs declaration using the LOF algorithm based on the weighted Euclidean distance, thereby achieving intelligent analysis of customs declaration data. This solution significantly improves the accuracy of customs declaration anomaly detection and the effectiveness of risk assessment, helping regulatory agencies to promptly identify potential risks, reduce regulatory costs, and improve overall risk control levels.

[0003] CN116542523A discloses a method and system for generating risk rules from customs declaration information, belonging to the field of data mining technology. This solution provides a method for generating risk rules from customs declaration information: S1, generating risk interval periods; S2, generating risk rules; S3, reviewing risk rules; S4, managing risk rules. A system for generating risk rules from customs declaration information includes a backend management module, a risk rule generation module, a risk rule review module, and a risk rule management module. The output of the backend management module is connected to the input of the risk rule generation module, the output of the risk rule generation module is connected to the input of the risk rule review module, and the output of the risk rule review module is connected to the input of the risk rule management module. This solution can meet the time and space requirements for risk rule generation under large-scale data, improving data mining efficiency.

[0004] Therefore, in related technologies, customs declaration data can be obtained based on customs declaration documents and the risks therein can be analyzed. However, it is not possible to classify and analyze the various types of information in the customs declaration documents to analyze different risk levels, thus failing to provide accurate risk detection results.

[0005] The information disclosed in the background section of this application is intended only to enhance the understanding of the general background of this application and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0006] This invention provides a method and system for customs clearance risk detection based on customs declaration document identification, which can solve the technical problem that related technologies cannot analyze different risk levels of customs declaration documents, thus failing to provide accurate risk detection results.

[0007] According to a first aspect of the present invention, a method for detecting customs clearance risks based on customs declaration document identification is provided, comprising:

[0008] Obtain scanned copies of several basic customs declaration documents and determine whether additional customs declaration documents are required;

[0009] If additional customs declaration documents are required, an upload prompt message will be generated, which will prompt the user to upload scanned copies of the additional customs declaration documents.

[0010] Based on the identification results of scanned copies of basic customs declaration documents and the upload results of scanned copies of additional customs declaration documents, the high customs declaration risk trigger result is determined.

[0011] Based on the identification results of scanned copies of basic customs declaration documents and scanned copies of additional customs declaration documents, the customs declaration risk triggering results are determined.

[0012] Based on the identification results of scanned copies of basic customs declaration documents, determine the low customs declaration risk trigger result;

[0013] The customs clearance risk detection results are determined based on the high customs declaration risk triggering results, medium customs declaration risk triggering results, and low customs declaration risk triggering results.

[0014] According to the present invention, based on the identification results of scanned copies of basic customs declaration documents and the upload results of scanned copies of additional customs declaration documents, a high customs declaration risk trigger result is determined, including:

[0015] If additional customs declaration documents are required, and scanned copies of the additional customs declaration documents are not received within the preset time period, the high customs declaration risk trigger result will be determined as triggering high risk.

[0016] Otherwise, based on the basic customs declaration documents, obtain the category information and first HS code information of the goods, and determine the first high customs declaration risk triggering condition;

[0017] Based on the basic customs declaration documents, obtain invoice amount information and contract amount information, and determine the second highest customs declaration risk triggering conditions;

[0018] Based on the basic customs declaration documents, obtain the first quantity information from the invoice, the second quantity information from the packing list, and the third quantity information from the bill of lading, and determine the third highest customs declaration risk triggering conditions;

[0019] If any one of the first, second, or third highest customs declaration risk triggering conditions is met, the high customs declaration risk triggering result will be determined as triggering high risk.

[0020] According to the present invention, based on basic customs declaration documents, the category information and first HS code information of the goods are obtained, and the first high customs declaration risk triggering condition is determined, including:

[0021] The HS coding classification model, after training, processes the category information of goods to obtain the predicted HS coding of the goods.

[0022] If the predicted HS code is different from the first HS code information, the historical classification information is queried according to the predicted HS code to determine that the HS code of the goods is an incorrect classification record of the predicted HS code.

[0023] In the error classification record, the percentage of errors in classifying the HS code of the goods as the first HS code information is determined;

[0024] Obtain the first tax rate from the first HS code information, and predict the second tax rate from the HS code information;

[0025] According to the formula

[0026]

[0027] Determine the triggering conditions for the highest customs declaration risk ,in, The second tax rate, The first tax rate, The percentage of errors, is the preset first probability threshold, and if is the conditional function.

[0028] According to the present invention, based on basic customs declaration documents, invoice amount information and contract amount information are obtained, and the second highest customs declaration risk triggering condition is determined, including:

[0029] Obtain information on the type of goods based on the basic customs declaration documents;

[0030] The category information is retrieved to determine the average price of the goods corresponding to the category information;

[0031] The average amount information is determined based on the average price and the first quantity information in the invoice;

[0032] According to the formula

[0033]

[0034] Determine the triggering conditions for the second highest customs declaration risk ,in, For invoice amount information, For contract amount information, For average amount information, is the preset second probability threshold, and if is the conditional function.

[0035] According to the present invention, based on basic customs declaration documents, first quantity information from the invoice, second quantity information from the packing list, and third quantity information from the bill of lading are obtained, and third high customs declaration risk triggering conditions are determined, including:

[0036] According to the formula

[0037]

[0038] Determine the triggering conditions for the third highest customs declaration risk ,in, For the first quantity information, For the second quantity information, This is the third quantity information. is the preset third probability threshold, if is the conditional function, and min is the minimum value function.

[0039] According to the present invention, based on the identification results of scanned copies of basic customs declaration documents and the identification results of scanned copies of additional customs declaration documents, the determination of customs declaration risk triggering results includes:

[0040] Based on the second and third quantity information, determine the first customs declaration risk triggering condition;

[0041] The validity period of the license is determined based on the identification results of the scanned copies of the additional customs declaration documents;

[0042] Based on the current date and license validity period, determine the second type of customs declaration risk triggering conditions;

[0043] If the predicted HS code differs from the first HS code information, then the third customs declaration risk triggering condition is determined based on the first tax rate, the second tax rate, and the error rate.

[0044] If any one of the first, second, or third customs declaration risk triggering conditions is met, the customs declaration risk triggering result will be determined as a medium risk.

[0045] According to the present invention, if the predicted HS code differs from the first HS code information, a third customs declaration risk triggering condition is determined based on the first tax rate, the second tax rate, and the error percentage, including:

[0046] According to the formula

[0047]

[0048] Determine the triggering conditions for third-party customs declaration risks ,in, For the preset ratio threshold, This is the preset fourth probability threshold.

[0049] According to the present invention, a low customs declaration risk trigger result is determined based on the identification result of scanned copies of basic customs declaration documents, including:

[0050] Based on the recognition results of the scanned copies of the basic customs declaration documents, the shipping addresses in various basic customs declaration documents are obtained;

[0051] In cases where the shipping addresses are inconsistent across multiple basic customs declaration documents, the low-risk declaration trigger result will be determined as triggering low risk.

[0052] According to a second aspect of the present invention, a customs clearance risk detection system based on customs declaration document identification is provided, comprising:

[0053] The judgment module is used to obtain scanned copies of multiple basic customs declaration documents and determine whether additional customs declaration documents are required.

[0054] The prompt module is used to generate upload prompt information if additional customs declaration documents are required. The upload prompt information is used to prompt the user to upload scanned copies of the additional customs declaration documents.

[0055] The High Customs Declaration Risk Trigger Result Module is used to determine the high customs declaration risk trigger result based on the recognition results of scanned copies of basic customs declaration documents and the upload results of scanned copies of additional customs declaration documents.

[0056] The customs declaration risk trigger result module is used to determine the customs declaration risk trigger result based on the recognition results of scanned copies of basic customs declaration documents and scanned copies of additional customs declaration documents.

[0057] The Low Customs Declaration Risk Trigger Result Module is used to determine the low customs declaration risk trigger result based on the recognition results of the scanned copies of basic customs declaration documents.

[0058] The customs clearance risk detection result module is used to determine the customs clearance risk detection result based on the high customs clearance risk trigger result, medium customs clearance risk trigger result, and low customs clearance risk trigger result.

[0059] By adopting the above technical solution, the present invention can achieve the following technical effects:

[0060] According to this invention, scanned copies of basic customs declaration documents can be collected, and various information can be obtained based on the recognition results. This information can then be classified and analyzed to determine the various levels of risk that the customs declaration documents may present. This allows for accurate alerts based on the risk level of the customs declaration documents, improving the accuracy of customs clearance risk detection and increasing the probability of successful customs clearance. When determining the trigger condition for the highest customs declaration risk, the probability of intentional misclassification can be represented by the error percentage, and the probability of tax evasion motive can be represented by the relative difference between the conditional function and the tax rate. Combining these two aspects comprehensively represents the probability of intentionally misclassifying goods into categories with lower tax rates, thereby determining the trigger condition for the highest customs declaration risk and accurately and objectively identifying high-risk situations involving suspected tax evasion. When determining the trigger condition for the second highest customs declaration risk, a conditional function can be used to compare invoice amount information, contract amount information, and average amount information, thereby accurately and objectively analyzing the suspicion and risk of tax evasion caused by price differences. When determining the third highest customs declaration risk trigger condition, a conditional function can be used to compare the first quantity information in the invoice, the second quantity information in the packing list, and the third quantity information in the bill of lading. This allows for an accurate and objective determination of whether there is any suspicion of underreporting of the quantity of goods or tax evasion. When determining the third medium customs declaration risk trigger condition, a conditional function can be used to determine the risk of property loss. Specifically, the conditional function can determine the probability of classification errors caused by staff negligence, thereby determining the probability of property loss due to staff negligence. This, in turn, determines the third medium customs declaration risk trigger condition, accurately and objectively representing the risk of property loss.

[0061] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only, and are not intended to limit the invention. Other features and aspects of the invention will become clearer from the following detailed description of exemplary embodiments with reference to the accompanying drawings. Attached Figure Description

[0062] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other embodiments can be obtained based on these drawings without creative effort.

[0063] Figure 1 A flowchart of a customs clearance risk detection method based on customs document identification according to an embodiment of the present invention is shown as an example;

[0064] Figure 2 An exemplary flowchart illustrating the determination of high customs declaration risk triggering results according to an embodiment of the present invention is shown;

[0065] Figure 3 An exemplary flowchart illustrating the determination of customs declaration risk triggering results according to an embodiment of the present invention is shown;

[0066] Figure 4 An exemplary flowchart illustrating the determination of low customs declaration risk triggering results according to an embodiment of the present invention is shown;

[0067] Figure 5 A block diagram of a customs clearance risk detection system based on customs document identification according to an embodiment of the present invention is shown as an example. Detailed Implementation

[0068] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0069] The technical solution of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0070] Figure 1 A flowchart of a customs clearance risk detection method based on customs document identification according to an embodiment of the present invention is shown as an example, the method comprising:

[0071] Step S1: Obtain scanned copies of multiple basic customs declaration documents and determine whether additional customs declaration documents are required.

[0072] Step S2: If additional customs declaration documents are required, an upload prompt message is generated. The upload prompt message is used to prompt the user to upload scanned copies of the additional customs declaration documents.

[0073] Step S3: Based on the recognition results of the scanned copies of the basic customs declaration documents and the upload results of the scanned copies of the additional customs declaration documents, determine the high customs declaration risk trigger result.

[0074] Step S4: Based on the identification results of the scanned copies of the basic customs declaration documents and the scanned copies of the additional customs declaration documents, determine the customs declaration risk trigger result.

[0075] Step S5: Based on the identification results of the scanned copies of the basic customs declaration documents, determine the low customs declaration risk trigger result;

[0076] Step S6: Determine the customs clearance risk detection result based on the high customs declaration risk triggering result, the medium customs declaration risk triggering result, and the low customs declaration risk triggering result.

[0077] According to an embodiment of the present invention, the customs clearance risk detection method based on customs declaration document recognition can collect scanned copies of basic customs declaration documents and obtain various information based on the recognition results. The various information can be classified and analyzed to determine the various levels of risk that the customs declaration documents may have. Thus, accurate prompts can be given based on the level of risk of the customs declaration documents, thereby improving the accuracy of customs clearance risk detection and the probability of smooth customs clearance.

[0078] According to one embodiment of the present invention, in step S1, the customs declaration documents may include various types of documents, such as basic customs declaration documents and additional customs declaration documents. Basic customs declaration documents may include invoices, packing lists, bills of lading, contracts, etc., while additional customs declaration documents may include licenses, quarantine certificates, insurance policies, etc. During the inspection of customs declaration documents, almost every batch of goods requires inspection of basic customs declaration documents, while only some special goods require inspection of additional customs declaration documents. Therefore, scanned copies of the basic customs declaration documents can be obtained first, and based on information such as the category of goods described in the basic customs declaration documents, it can be determined whether additional customs declaration documents need to be inspected. For example, if the imported goods are products made from animals or plants, a quarantine certificate may need to be inspected.

[0079] According to one embodiment of the present invention, in step S2, if additional customs declaration documents need to be verified, an upload prompt message is generated, prompting the user to upload scanned copies of the additional customs declaration documents. For example, the prompt message may include the type of additional customs declaration documents to be uploaded.

[0080] According to one embodiment of the present invention, in step S3, a high customs declaration risk trigger result can be determined based on the identification result of the scanned copies of the basic customs declaration documents and the upload result of the scanned copies of the additional customs declaration documents. The high customs declaration risk trigger result can be used to indicate whether there is a high risk associated with the customs declaration documents, such as significant risks like tax evasion or incomplete customs declaration documents. Once such a risk-related event occurs, the probability of customs clearance failure is very high, and it may trigger other risks, such as detention of goods, return of goods, fines, or even legal liability.

[0081] Figure 2 An exemplary flowchart illustrating the determination of high customs declaration risk triggering results according to an embodiment of the present invention is shown.

[0082] According to an embodiment of the present invention, step S3 includes: step S31, if additional customs declaration documents are required and no scanned copies of the additional customs declaration documents are received within a preset time period, then the high customs declaration risk triggering result is determined as triggering high risk; step S32, otherwise, based on the basic customs declaration documents, the category information and the first HS code information of the goods are obtained, and the first high customs declaration risk triggering condition is determined; step S33, based on the basic customs declaration documents, the invoice amount information and the contract amount information are obtained, and the second high customs declaration risk triggering condition is determined; step S34, based on the basic customs declaration documents, the first quantity information in the invoice, the second quantity information in the packing list, and the third quantity information in the bill of lading are obtained, and the third high customs declaration risk triggering condition is determined; step S35, if any one of the first, second, and third high customs declaration risk triggering conditions is met, the high customs declaration risk triggering result is determined as triggering high risk.

[0083] According to an embodiment of the present invention, in step S31, the preset time period can be set to 24 hours. If additional customs declaration documents are required, but scanned copies of the additional customs declaration documents cannot be uploaded within the preset time period, then the high customs declaration risk trigger result is determined as triggering high risk, that is, there is a risk that the documents are incomplete and customs clearance is impossible.

[0084] According to one embodiment of the present invention, in step S32, if no additional customs declaration documents are required, or if scanned copies of additional customs declaration documents are uploaded within a preset time period, then the presence of high risk is determined from other aspects.

[0085] According to an embodiment of the present invention, step S32 includes: step S321, processing the category information of goods through the trained HS code classification model to obtain the predicted HS code of the goods; step S322, if the predicted HS code is different from the first HS code information, querying the historical classification information according to the predicted HS code to determine the HS code of the goods as an incorrect classification record of the predicted HS code; step S323, determining the percentage of incorrect classification records in which the HS code of the goods is classified as the first HS code information; step S324, obtaining the first tax rate of the first HS code information and the second tax rate of the predicted HS code information; step S325, determining the first high customs declaration risk triggering condition according to formula (1). ,

[0086] (1)

[0087] in, The second tax rate, The first tax rate, The percentage of errors, is the preset first probability threshold, and if is the conditional function.

[0088] According to one embodiment of the present invention, the HS code is the customs code. In basic customs declaration documents (e.g., contracts), the category information of goods can be recorded. The HS code classification model can be a neural network model. The HS code classification model can be trained before use. For example, the category information of goods can be input into the HS code classification model, and the HS code classification model can output the HS code. The output HS code can be compared with the actual HS code of the goods to determine the error between the two. This error can be used as the loss function of the HS code classification model. The loss function can be backpropagated to adjust the parameters of the HS code classification model to minimize the loss function. After repeated training, the trained HS code classification model can be obtained.

[0089] According to one embodiment of the present invention, a trained HS code classification model is used to process the category information of goods to obtain a predicted HS code for the goods. This predicted HS code can be used as the accurate HS code for verification against the first HS code in the basic customs declaration documents. If the predicted HS code differs from the first HS code, the cause of the error can be analyzed. For example, the possibility of tax evasion can be analyzed, thereby analyzing the potential risks arising from the discrepancy between the predicted HS code and the first HS code.

[0090] According to one embodiment of the present invention, historical classification information of multiple batches of goods with predicted HS codes can be queried in a database. If there are misclassified goods, the misclassification records of these goods are obtained, that is, which categories these goods were misclassified into and which HS codes these categories have. The proportion of goods misclassified into the category corresponding to the first HS code information can be determined, i.e., the error percentage. If the error percentage is high, it indicates that the category corresponding to the first HS code is very similar to the category corresponding to the predicted HS code, meaning they are easily confused during classification and prone to errors during manual classification. Conversely, if the error percentage is low, the category corresponding to the first HS code is not easily confused with the category corresponding to the predicted HS code, and there may be suspicion of intentional misclassification, for example, intentionally misclassifying to a category with a lower tax rate.

[0091] According to one embodiment of the present invention, a first tax rate for the first HS code information and a second tax rate for the predicted HS code information can be obtained. That is, the first tax rate for the goods category corresponding to the first HS code and the second tax rate for the goods category corresponding to the predicted HS code. The first and second tax rates can be used to comprehensively determine whether there is a risk of tax evasion.

[0092] According to an embodiment of the present invention, in formula (1), Let be a conditional function, representing that In this case, the value of the conditional function is Otherwise, the conditional function value is 0. If the second tax rate is higher than the first tax rate, there may be an incentive or tendency to misclassify the HS code of goods as the first HS code information. The relative deviation between the first and second tax rates can indicate the likelihood of an incentive to evade taxes. Conversely, if... Therefore, there is no motive for tax evasion.

[0093] According to an embodiment of the present invention, as described above, if the error rate is high (e.g., exceeding 50%), it indicates that the category corresponding to the first HS code is quite similar to the category corresponding to the predicted HS code, that is, it is easy to confuse during classification and easy to make mistakes during manual classification. Therefore, the possibility of intentional misclassification is low. Thus, the above error rate can represent the possibility of unintentional misclassification, and can be used... This indicates the likelihood of intentional misclassification. Conversely, if the error rate is low, the category corresponding to the first HS code and the category corresponding to the predicted HS code are not easily confused, suggesting a possible intentional misclassification. In cases where confusion is unlikely, the probability of unintentional misclassification remains high. The possibility of intentional misclassification remains. .

[0094] According to one embodiment of the present invention, multiplying the above two items yields the possibility of intentionally misclassifying a product into a category with a lower tax rate. That is, the possibility of misclassification due to intent to evade taxes. If this possibility is higher than a preset first probability threshold (e.g., 0.2), then the first high customs declaration risk trigger condition is determined to be met.

[0095] In this way, the probability of intentional misclassification can be represented by the error rate, and the probability of tax evasion can be represented by the relative difference between the conditional function and the tax rate. By combining the two aspects, the probability of intentionally misclassifying to a category with a lower tax rate can be comprehensively represented, thereby determining the highest customs declaration risk triggering condition, so as to accurately and objectively determine whether there is a high-risk situation suspected of tax evasion.

[0096] According to an embodiment of the present invention, in step S33, the existence of a high-risk situation suspected of tax evasion can be determined from the perspective of the amount involved in the goods. Step S33 includes: step S331, obtaining the category information of the goods based on the basic customs declaration documents; step S332, retrieving the category information to determine the average price of the goods corresponding to the category information; step S333, determining the average amount information based on the average price and the first quantity information in the invoice; step S334, determining the second high customs declaration risk triggering condition according to formula (2). ,

[0097] (2)

[0098] in, For invoice amount information, For contract amount information, For average amount information, is the preset second probability threshold, and if is the conditional function.

[0099] According to one embodiment of the present invention, category information can be obtained from basic customs declaration documents and retrieved from multiple online trading platforms to determine the average price of goods corresponding to that category information, thereby determining whether there is a possibility of deliberately lowering the price of goods transported in this shipment to evade taxes. The average price can be multiplied by first quantity information to obtain average amount information, that is, the amount required to purchase the same quantity of goods from the online trading platform, and the average amount information can be compared with the amounts in the invoice and contract.

[0100] According to an embodiment of the present invention, in formula (2), Let be a conditional function, representing that Under the given condition, the condition function value is 1; otherwise, the condition function value is equal to the inner condition function. The function value. Since the invoice is the direct proof of tax collection, if the invoice amount is less than the contract amount, the conditional function value is directly set to 1, that is, it is directly judged that there is a possibility of tax evasion. Conversely, if the invoice amount is greater than or equal to the contract amount, the inner conditional function can continue to judge whether the contract amount is reasonable. In the inner conditional function, if Then the value of the inner conditional function is Otherwise, the value is 0. That is, if the contract amount is lower than the average amount, there is a possibility of deliberately lowering the contract price to reduce taxes, i.e., there is a possibility of tax evasion, and this possibility can be determined. That is, the relative difference between the average amount information and the contract amount information. The higher the relative difference, the greater the reduction in tax and the higher the possibility of tax evasion. Conversely, if the contract amount information is greater than or equal to the average amount information, there is no possibility of tax evasion, and the value of the inner condition function is directly set to 0.

[0101] According to an embodiment of the present invention, based on the above analysis, the conditional function value in formula (2) can represent the possibility of suspected tax evasion from the perspective of amount. If the possibility is greater than or equal to a preset second probability threshold (e.g., 30%), then a high-risk situation of suspected tax evasion is determined.

[0102] In this way, conditional functions can be used to compare invoice amount information, contract amount information, and average amount information, thereby accurately and objectively analyzing the suspicion and risk of tax evasion caused by price differences.

[0103] According to an embodiment of the present invention, in step S34, the possibility of high-risk situations involving suspected tax evasion can be analyzed from the perspective of the quantity of goods. Step S34 includes: obtaining first quantity information from the invoice, second quantity information from the packing list, and third quantity information from the bill of lading based on basic customs declaration documents, and determining the third high customs declaration risk triggering condition, including: determining the third high customs declaration risk triggering condition according to formula (3). ,

[0104] (3)

[0105] in, For the first quantity information, For the second quantity information, This is the third quantity information. is the preset third probability threshold, if is the conditional function, and min is the minimum value function.

[0106] According to one embodiment of the present invention, in formula (3), For a conditional function, if If the first quantity information is less than or equal to the minimum of the third quantity information, the quantity of goods on the invoice is suspected of being underreported, potentially leading to tax evasion. Therefore, in this case, the conditional function value is set to 1; otherwise, there is no suspicion of underreporting. The third probability threshold can be set to 0.5. If the conditional function value is 1, the third highest customs declaration risk triggering condition is met; if the conditional function value is 0, the third highest customs declaration risk triggering condition is not met.

[0107] In this way, conditional functions can be used to compare the first quantity information in the invoice, the second quantity information in the packing list, and the third quantity information in the bill of lading, thereby accurately and objectively determining whether there is any suspicion of underreporting of the quantity of goods or tax evasion.

[0108] According to one embodiment of the present invention, in step S35, if any one of the first, second, and third high customs declaration risk triggering conditions is met, the high customs declaration risk triggering result is determined to be a high risk trigger. If no additional customs declaration documents are required, or if scanned copies of additional customs declaration documents can be uploaded within a preset time period, and none of the first, second, and third high customs declaration risk triggering conditions are met, then it can be determined that no high risk is triggered.

[0109] According to an embodiment of the present invention, in step S4, the result of triggering customs declaration risk can be determined based on the identification results of the scanned copies of the basic customs declaration documents and the scanned copies of the additional customs declaration documents. For example, there may be inconsistencies in the basic customs declaration documents or risks such as the expiration of the licenses for the additional customs declaration documents. If these exist, it may be difficult to clear customs.

[0110] Figure 3 An exemplary flowchart illustrating the determination of customs declaration risk triggering results according to an embodiment of the present invention is shown.

[0111] According to an embodiment of the present invention, step S4 includes: step S41, determining a first medium-level customs declaration risk triggering condition based on the second quantity information and the third quantity information; step S42, determining the license validity period based on the identification result of the scanned copy of the additional customs declaration document; step S43, determining a second medium-level customs declaration risk triggering condition based on the current date and the license validity period; step S44, if the predicted HS code is different from the first HS code information, determining a third medium-level customs declaration risk triggering condition based on the first tax rate, the second tax rate, and the error ratio; step S45, if any one of the first medium-level customs declaration risk triggering condition, the second medium-level customs declaration risk triggering condition, and the third medium-level customs declaration risk triggering condition is met, determining the medium-level customs declaration risk triggering result as triggering medium risk.

[0112] According to an embodiment of the present invention, in step S41, if the second quantity information and the third quantity information are inconsistent, that is, the quantity of goods in the packing list and the quantity of goods in the bill of lading are inconsistent, it is difficult to explain the actual quantity of goods, making it difficult to clear customs. Therefore, in the case of inconsistency between the second quantity information and the third quantity information, it is determined that the first customs declaration risk triggering condition is met.

[0113] According to one embodiment of the present invention, in step S42, the validity period of the license for the additional customs declaration documents can be identified and compared with the current date to determine whether the additional customs declaration documents have expired. If they have expired, it will be difficult to pass through customs, and it may be necessary to reissue them before customs clearance can proceed. That is, if the current date has exceeded the license validity period, the second customs declaration risk triggering condition is met.

[0114] According to one embodiment of the present invention, in step S43, it can be determined whether the difference between the predicted HS code and the first HS code information is due to staff oversight, and not due to suspected tax evasion.

[0115] According to an embodiment of the present invention, step S43 includes: determining the third customs declaration risk triggering condition according to formula (4). ,

[0116] (4)

[0117] in, For the preset ratio threshold, This is the preset fourth probability threshold.

[0118] According to one embodiment of the present invention, in formula (4), For a conditional function, in In the given case, the condition function evaluates to 0; otherwise, it evaluates to 0. As mentioned above, if If the predicted HS code differs from the first HS code, there is suspicion of tax evasion, which may meet the first high customs declaration risk triggering condition, but not the third customs declaration risk triggering condition. Therefore, the condition function can be set to 0. In other words, misclassifying a product into a category corresponding to the higher-tax-rate HS code carries the risk of financial loss, the magnitude of which is the relative difference between the first and second tax rates. .

[0119] According to one embodiment of the present invention, in formula (4), For a conditional function, if Then the value of the conditional function is Otherwise If the error rate is higher than or equal to a preset threshold (e.g., 0.5), the error rate is high, indicating that the predicted HS code and the category corresponding to the first HS code information are easily confused. The error rate can represent the probability that easily confused categories are confused due to staff negligence, or the probability of property loss due to staff negligence. If the error rate is lower than the preset threshold, it indicates that the predicted HS code and the category corresponding to the first HS code information are not easily confused, but the staff made a misclassification. Furthermore, in the third customs declaration risk triggering condition, the staff's misclassification was not intentional, but rather a classification error due to the staff's lack of understanding of the classification rules. This can be used as a measure of the probability that staff will confuse the predicted HS code with the category corresponding to the first HS code information. This can be used as the probability that a worker's lack of understanding of the classification rules leads to a classification error. Therefore, this conditional function can be used to represent the probability that a worker's negligence or lack of understanding of the classification rules causes a classification error; that is, the probability that a classification error is caused by the worker's negligence.

[0120] According to one embodiment of the present invention, the above two conditional functions can be multiplied to obtain the probability of property loss risk caused by staff. If the probability is higher than or equal to a preset fourth probability threshold (e.g., 0.5), then the third customs declaration risk triggering condition is met.

[0121] In this way, the risk of property loss can be determined through conditional functions, and the probability of classification errors caused by staff can be determined through conditional functions, thereby determining the probability of property loss risk caused by staff, and then determining the third customs declaration risk triggering condition, so as to accurately and objectively represent the risk of property loss.

[0122] According to one embodiment of the present invention, in step S45, if any one of the first, second, and third customs declaration risk triggering conditions is met, the customs declaration risk triggering result is determined to be a medium risk, that is, there is a possibility that customs clearance cannot be completed. If none of the first, second, and third customs declaration risk triggering conditions are met, then a medium risk is not triggered.

[0123] According to an embodiment of the present invention, in step S5, a low customs declaration risk trigger result can be determined based on the identification result of the scanned copy of the basic customs declaration document. For example, based on the identification result of the scanned copy of the customs declaration document, it can be determined whether some less important information is filled in correctly. If it is not filled in correctly, a low risk is triggered, and it needs to be modified before it can be cleared through customs.

[0124] Figure 4 An exemplary flowchart illustrating the determination of low customs declaration risk triggering results according to an embodiment of the present invention is shown.

[0125] According to an embodiment of the present invention, step S5 includes: step S51, obtaining the shipping address in multiple basic customs declaration documents based on the identification result of the scanned copy of the basic customs declaration document; step S52, in the case that the shipping addresses in multiple basic customs declaration documents are inconsistent, determining the low customs declaration risk triggering result as triggering low risk.

[0126] According to one embodiment of the present invention, in addition to the above-mentioned inaccurate shipping address, it is also possible to verify whether the consignee's name, contact information, etc. are accurate, that is, whether the above information is consistent in various basic customs declaration documents. If they are inconsistent, a low-risk condition is triggered, and modifications are required before customs clearance.

[0127] According to one embodiment of the present invention, in step S6, the level of risk can be determined based on the high, medium, and low customs declaration risk triggering results to determine the customs clearance risk detection result. If the high, medium, and low customs declaration risk triggering results are all untriggered, it indicates that no customs clearance risk has been detected, and the probability of successful customs clearance is relatively high.

[0128] According to embodiments of the present invention, a customs clearance risk detection method based on customs declaration document identification can collect scanned copies of basic customs declaration documents and obtain various information based on the identification results. This information is then classified and analyzed to determine the potential risk levels associated with the customs declaration documents. This allows for accurate alerts based on the risk level, improving the accuracy of customs clearance risk detection and increasing the probability of successful customs clearance. When determining the first highest customs clearance risk trigger condition, the probability of intentional misclassification can be represented by the error percentage, and the relative difference between the conditional function and the tax rate can represent the potential motivation for tax evasion. Combining these two aspects comprehensively indicates the possibility of intentionally misclassifying goods to categories with lower tax rates, thereby determining the first highest customs clearance risk trigger condition and accurately and objectively identifying high-risk situations involving suspected tax evasion. When determining the second highest customs clearance risk trigger condition, a conditional function can be used to compare invoice amount information, contract amount information, and average amount information, thereby accurately and objectively analyzing the suspicion and risk of tax evasion caused by price differences. When determining the third highest customs declaration risk trigger condition, a conditional function can be used to compare the first quantity information in the invoice, the second quantity information in the packing list, and the third quantity information in the bill of lading. This allows for an accurate and objective determination of whether there is any suspicion of underreporting of the quantity of goods or tax evasion. When determining the third medium customs declaration risk trigger condition, a conditional function can be used to determine the risk of property loss. Specifically, the conditional function can determine the probability of classification errors caused by staff negligence, thereby determining the probability of property loss due to staff negligence. This, in turn, determines the third medium customs declaration risk trigger condition, accurately and objectively representing the risk of property loss.

[0129] Figure 5 An exemplary block diagram of a customs clearance risk detection system based on customs declaration document identification according to an embodiment of the present invention is shown, the system comprising:

[0130] The judgment module is used to obtain scanned copies of multiple basic customs declaration documents and determine whether additional customs declaration documents are required.

[0131] The prompt module is used to generate upload prompt information if additional customs declaration documents are required. The upload prompt information is used to prompt the user to upload scanned copies of the additional customs declaration documents.

[0132] The High Customs Declaration Risk Trigger Result Module is used to determine the high customs declaration risk trigger result based on the recognition results of scanned copies of basic customs declaration documents and the upload results of scanned copies of additional customs declaration documents.

[0133] The customs declaration risk trigger result module is used to determine the customs declaration risk trigger result based on the recognition results of scanned copies of basic customs declaration documents and scanned copies of additional customs declaration documents.

[0134] The Low Customs Declaration Risk Trigger Result Module is used to determine the low customs declaration risk trigger result based on the recognition results of the scanned copies of basic customs declaration documents.

[0135] The customs clearance risk detection result module is used to determine the customs clearance risk detection result based on the high customs clearance risk trigger result, medium customs clearance risk trigger result, and low customs clearance risk trigger result.

[0136] This invention can be a method, apparatus, system, and / or computer program product. The computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for performing various aspects of the invention.

[0137] Those skilled in the art should understand that the embodiments of the present invention described above and shown in the accompanying drawings are merely examples and do not limit the present invention. The objectives of the present invention have been fully and effectively achieved. The functions and structural principles of the present invention have been demonstrated and explained in the embodiments, and any variations or modifications may be made to the implementation of the present invention without departing from the stated principles.

[0138] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for detecting customs clearance risks based on customs declaration document identification, characterized in that, include: Obtain scanned copies of several basic customs declaration documents and determine whether additional customs declaration documents are required; If additional customs declaration documents are required, an upload prompt message will be generated, which will prompt the user to upload scanned copies of the additional customs declaration documents. Based on the identification results of scanned copies of basic customs declaration documents and the upload results of scanned copies of additional customs declaration documents, the high customs declaration risk trigger result is determined. Based on the identification results of scanned copies of basic customs declaration documents and scanned copies of additional customs declaration documents, the customs declaration risk triggering results are determined. Based on the identification results of scanned copies of basic customs declaration documents, determine the low customs declaration risk trigger result; The customs clearance risk detection results are determined based on the high customs declaration risk triggering results, medium customs declaration risk triggering results, and low customs declaration risk triggering results. Based on the identification results of scanned copies of basic customs declaration documents and the upload results of scanned copies of additional customs declaration documents, high customs declaration risk triggering results are determined, including: If additional customs declaration documents are required, and scanned copies of the additional customs declaration documents are not received within the preset time period, the high customs declaration risk trigger result will be determined as triggering high risk. Otherwise, based on the basic customs declaration documents, obtain the category information and first HS code information of the goods, and determine the first high customs declaration risk triggering condition; Based on the basic customs declaration documents, obtain invoice amount information and contract amount information, and determine the second highest customs declaration risk triggering conditions; Based on the basic customs declaration documents, obtain the first quantity information from the invoice, the second quantity information from the packing list, and the third quantity information from the bill of lading, and determine the third highest customs declaration risk triggering conditions; If any one of the first, second, or third highest customs declaration risk triggering conditions is met, the result of triggering high customs declaration risk will be determined as triggering high risk. Based on the basic customs declaration documents, obtain the category information and first HS code information of the goods, and determine the first high customs declaration risk triggering conditions, including: The HS coding classification model, after training, processes the category information of goods to obtain the predicted HS coding of the goods. If the predicted HS code is different from the first HS code information, the historical classification information is queried according to the predicted HS code to determine that the HS code of the goods is an incorrect classification record of the predicted HS code. In the error classification record, the percentage of errors in classifying the HS code of the goods as the first HS code information is determined; Obtain the first tax rate from the first HS code information, and predict the second tax rate from the HS code information; According to the formula Determine the triggering conditions for the highest level of customs declaration risk. ,in, The second tax rate, The first tax rate, The percentage of errors, is the preset first probability threshold, and if is the conditional function.

2. The customs clearance risk detection method based on customs declaration document identification according to claim 1, characterized in that, Based on the basic customs declaration documents, obtain invoice amount information and contract amount information, and determine the second highest customs declaration risk triggering conditions, including: Obtain information on the type of goods based on the basic customs declaration documents; The category information is retrieved to determine the average price of the goods corresponding to the category information; The average amount information is determined based on the average price and the first quantity information in the invoice; According to the formula Determine the triggering conditions for the second highest customs declaration risk ,in, For invoice amount information, For contract amount information, For average amount information, is the preset second probability threshold, and if is the conditional function.

3. The customs clearance risk detection method based on customs declaration document identification according to claim 1, characterized in that, Based on the basic customs declaration documents, obtain the first quantity information from the invoice, the second quantity information from the packing list, and the third quantity information from the bill of lading, and determine the third highest customs declaration risk triggering conditions, including: According to the formula Determine the triggering conditions for the third highest customs declaration risk ,in, For the first quantity information, For the second quantity information, This is the third quantity information. is the preset third probability threshold, if is the conditional function, and min is the minimum value function.

4. The customs clearance risk detection method based on customs declaration document identification according to claim 1, characterized in that, Based on the identification results of scanned copies of basic customs declaration documents and scanned copies of additional customs declaration documents, the triggering results for customs declaration risks are determined, including: Based on the second and third quantity information, determine the first customs declaration risk triggering condition; The validity period of the license is determined based on the identification results of the scanned copies of the additional customs declaration documents; Based on the current date and license validity period, determine the second type of customs declaration risk triggering conditions; If the predicted HS code differs from the first HS code information, then the third customs declaration risk triggering condition is determined based on the first tax rate, the second tax rate, and the error rate. If any one of the first, second, or third customs declaration risk triggering conditions is met, the customs declaration risk triggering result will be determined as a triggered medium risk.

5. The customs clearance risk detection method based on customs declaration document identification according to claim 4, characterized in that, If the predicted HS code differs from the first HS code information, then based on the first tariff rate, the second tariff rate, and the error percentage, the third customs declaration risk triggering condition is determined, including: According to the formula Determine the triggering conditions for third-party customs declaration risks ,in, For the preset ratio threshold, This is the preset fourth probability threshold.

6. The customs clearance risk detection method based on customs declaration document identification according to claim 1, characterized in that, Based on the identification results of scanned copies of basic customs declaration documents, low customs declaration risk triggering results are determined, including: Based on the recognition results of the scanned copies of the basic customs declaration documents, the shipping addresses in various basic customs declaration documents are obtained; In cases where the shipping addresses are inconsistent across multiple basic customs declaration documents, the low-risk declaration trigger result will be determined as triggering low risk.

7. A customs clearance risk detection system based on customs declaration document recognition, used to execute the method as described in any one of claims 1-6, characterized in that, include: The judgment module is used to obtain scanned copies of multiple basic customs declaration documents and determine whether additional customs declaration documents are required. The prompt module is used to generate upload prompt information if additional customs declaration documents are required. The upload prompt information is used to prompt the user to upload scanned copies of the additional customs declaration documents. The High Customs Declaration Risk Trigger Result Module is used to determine the high customs declaration risk trigger result based on the recognition results of scanned copies of basic customs declaration documents and the upload results of scanned copies of additional customs declaration documents. The customs declaration risk trigger result module is used to determine the customs declaration risk trigger result based on the recognition results of scanned copies of basic customs declaration documents and scanned copies of additional customs declaration documents. The Low Customs Declaration Risk Trigger Result Module is used to determine the low customs declaration risk trigger result based on the recognition results of the scanned copies of basic customs declaration documents. The customs clearance risk detection result module is used to determine the customs clearance risk detection result based on the high customs clearance risk trigger result, medium customs clearance risk trigger result, and low customs clearance risk trigger result.

Images