Intelligent auditing method for cargo approval report

Abstract

The invention relates to the field of data analysis and auditing, in particular to an intelligent auditing method for a cargo approval report, and the method comprises the steps: obtaining the form data of the cargo approval report received by a receiving end, positioning a keyword in the form data, and recognizing a serial number corresponding to the keyword; calling a serial number matching feature corresponding to a cargo approval report submitted by a single client at a time so as to calculate a submission error representation value of the single cargo approval report, and dividing error abnormal degree categories of the cargo approval report; and adaptively checking and analyzing the cargo approval report according to the error abnormity degree category. According to the method and the device, the certificate number error risk is managed and controlled in advance, the subsequent transportation link problem is reduced, the accuracy required by certificate number auditing of aviation consignment and the auditing efficiency of aviation consignment are improved, and the transportation risk is reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of data analysis auditing, in particular to an intelligent auditing method for cargo approval report. BACKGROUND

[0002] The process of air shipment of goods involves multiple subjects such as the shipper, the carrier airline, airport security, customs, etc., and the types of goods are various, including ordinary goods, special goods such as dangerous goods, fresh food, etc., and the goods certificate number is the "information key" to connect each link. The aviation transportation has very high requirements for safety, timeliness and compliance. If the goods certificate number is incorrect, for example, the number is missing, miswritten, or does not match the actual goods, it may cause serious consequences, for example, causing problems in subsequent customs clearance, security check, sorting, etc., for example, the code does not match the goods, resulting in the goods being detained. In addition, air shipment of goods is time-sensitive, and the speed of certificate code auditing directly affects the turnover efficiency of goods loading and transportation. Therefore, by verifying the risk in advance, the goods certificate number is audited and verified to ensure the accuracy of the number audit while taking into account the efficiency of the audit.

[0003] Chinese patent application publication No. CN119850354A discloses an intelligent single auditing method, system, device and medium based on artificial intelligence. The method comprises: obtaining unstructured business data; performing information extraction and transformation on the unstructured business data based on a pre-trained small model to obtain structured text data; identifying pre-prepared approval system documents to generate approval rules and building a rule engine based on the approval rules; performing preliminary compliance audit on the structured text data based on the rule engine to obtain a first audit result; finding out associated knowledge features associated with the structured text data; performing semantic analysis on the structured text data based on a pre-trained large model to obtain semantic features; comparing and auditing the associated knowledge features and the semantic features to obtain a second audit result; and generating a single report based on the first audit result and the second audit result. The application improves the accuracy and consistency of the audit.

[0004] However, the existing technology still has the following problems, When auditing the certificate number of the goods carried by the aviation, the subjective judgment of "determining whether the error is acceptable based on experience" used by traditional manual auditing is easy to ignore the subtle errors in the number, which reduces the accuracy of the audit. In addition, air shipment has very high requirements for timeliness, and the efficiency of the audit directly affects whether the goods can be loaded on time. The number of goods carried by the aviation is large, and a large number of certificate numbers need to be audited. Manual auditing greatly reduces the efficiency of the audit. SUMMARY

[0005] To this end, the application provides an intelligent auditing method for cargo approval reports to overcome the problem that manual auditing is prone to overlook minor errors in numbers, reducing the accuracy of auditing, and the efficiency of manually auditing a large number of certificate numbers is low.

[0006] To achieve the above-mentioned purpose, the application provides an intelligent auditing method for cargo approval reports, which comprises: obtaining form data of a cargo approval report received by a receiving end, locating a keyword in the form data, and identifying a number corresponding to the keyword; calling a number matching feature corresponding to a cargo approval report applied by a single client in a single time to calculate an application error representation value of a single cargo approval report, and dividing the cargo approval report into an error abnormality degree category; According to the error abnormality degree category, the cargo approval report is analyzed, including, comparing the certificate number with a compliance number in a compliance number library, extracting a number error feature of the shipped cargo, determining an approval deviation value of the certificate number, and evaluating whether it meets the error allowed benchmark to process the cargo approval report; or, obtaining the proportion of the application error report in the cargo approval report applied by the client in a single time, determining whether to mark the client, and marking the application error report as a doubtful report; The number matching feature includes the number missing proportion and the number coincidence degree, and the number error feature includes the spatial distance between error characters and compliance characters and the minimum adjacent distance between error characters in the certificate number.

[0007] Further, the process of calculating the application error representation value of a single cargo approval report comprises: the ratio of the number missing proportion to the number missing proportion threshold value as the first application error feature; the ratio of the number coincidence degree threshold value to the number coincidence degree as the second application error feature; the sum of the first application error feature and the second application error feature as the application error representation value.

[0008] Further, the error abnormality degree category of the cargo approval report is divided, comprising: If the application error representation value of the cargo approval report is greater than or equal to the application error representation threshold value, the cargo approval report is divided into a high error abnormality degree category; If the application error representation value of the cargo approval report is less than the application error representation threshold value, the cargo approval report is divided into a low error abnormality degree category.

[0009] Further, according to the error abnormality degree category, the cargo approval report is audited and analyzed, including: If the cargo approval report is of a low error abnormality degree category, the certificate number is compared with the compliance number in the compliance number library, the number error characteristics of the shipped cargo are extracted, the approval deviation value of the certificate number is determined, and it is evaluated whether the error allowance benchmark is met, and the cargo approval report is processed; If the cargo approval report is of a high error abnormality degree category, the proportion of error report in the cargo approval report of the client single application is obtained, it is determined whether to mark the client, and the error report is marked as a suspicious report.

[0010] Further, the process of determining the approval deviation value of the certificate number includes: The spatial deviation value of the spatial distance between the error characters and the compliance characters and the spatial distance threshold value is calculated; The adjacency deviation value of the minimum adjacency distance between the error characters in the certificate number and the minimum adjacency distance threshold value is calculated; The sum of the spatial deviation value and the adjacency deviation value is taken as the approval deviation value.

[0011] Further, the evaluation of whether the error allowance benchmark is met includes: If the approval deviation value of the certificate number is greater than or equal to the approval deviation threshold value, the certificate number is evaluated as not meeting the error allowance benchmark; If the approval deviation value of the certificate number is less than the approval deviation threshold value, the certificate number is evaluated as meeting the error allowance benchmark.

[0012] Further, the cargo approval report is processed, including: If the certificate number meets the error allowance benchmark, the corresponding compliance number is marked, the picture of the shipped cargo corresponding to the certificate number is called, the actual number of the shipped cargo is analyzed, the number alignment degree between the actual number and the compliance number is determined, and the corresponding cargo approval report is processed and marked; If the certificate number does not meet the error allowance benchmark, the cargo approval report corresponding to the certificate number is marked as a suspicious report.

[0013] Further, the process of determining the number alignment degree between the actual number and the compliance number includes: The character corresponding number ratio of the actual number and the compliance number is taken as the number alignment degree.

[0014] Further, the corresponding cargo approval report is processed and marked, including: If the number alignment degree of the actual number and the compliance number is greater than or equal to the number alignment degree threshold, it is determined that the corresponding cargo approval report can be modified; If the number alignment degree of the actual number and the compliance number is less than the number alignment degree threshold, the corresponding cargo approval report is marked as a suspicious report.

[0015] Further, it is determined whether to mark the client, comprising: If the proportion of error reports is greater than or equal to the proportion threshold, the client is marked; Wherein, the cargo approval report with a report error representation value greater than or equal to a report error representation threshold is determined as the error report.

[0016] Compared with the prior art, the application obtains form data of the cargo approval report received by the receiving end, locates the keywords in the form data, and identifies the numbers corresponding to the keywords; the number matching features corresponding to the cargo approval report submitted by a single client at a time are called to calculate the report error representation value of a single cargo approval report, and the error abnormality degree category of the cargo approval report is divided; according to the error abnormality degree category, the cargo approval report is adaptively audited and analyzed, the certificate number error risk is pre-controlled, the subsequent transportation link problems are reduced, and the accuracy required by the aviation consignment for the certificate number audit and the aviation consignment audit efficiency are improved, and the transportation risk is reduced.

[0017] Especially, the application specifically analyzes the report submitted by a single client at a time, rather than mixed cargo approval reports submitted by multiple clients or a single client multiple times, instantly identifies high-risk reports submitted at a time, ensures that the risk at this time is locked in the first time, and avoids the risk lag caused by waiting for “multiple data accumulation”.

[0018] Especially, the application considers the influence of the error of the cargo certificate number on the transportation compliance in the air shipment, aggregates the "integrity defect" through the proportion of number missing, directly reflects the proportion of the number of missing characters in the total number in the number, can intuitively distinguish "partial missing" and "large area missing", and quantifies the possible damage degree of the core identification function of the number; focus on the "matching degree defect" through the number coincidence degree, reflect the matching proportion of the same characters in the certificate number and the compliance number, can judge whether the error is "local error" or "overall error", the combination of the two can convert the fuzzy judgment of "number error abnormality degree" into quantifiable numerical characteristics, ensure the consistency and accuracy of the number error evaluation, and then, the calculation of the error characteristic value provides data support for the subsequent division of the error abnormality degree category of the cargo approval report, the application controls the risk of certificate number error in advance, reduces the subsequent transportation problems, and improves the accuracy required for the certificate number audit of air shipment and the audit efficiency of air shipment, reduces the transportation risk.

[0019] Especially, for the fine audit of the cargo approval report of the low error abnormality degree category, avoid putting the correct number of goods into the suspicious report label, the application quantifies the randomness of the input error by the spatial distance between the error character and its corresponding compliance character, avoids the excessive interception of "non-subjective error"; combined with the analysis of the minimum adjacent distance between the error characters in the certificate number to judge the relevance of the error, distinguish whether the formed error is "concentrated error" or "dispersion error", to evaluate the risk degree, therefore, the application determines the approval deviation value of the certificate number by considering the above situation, quantifies the severity of the error itself, and indirectly evaluates the potential risk of the error, provides data support for subsequent evaluation whether to meet the error allowed benchmark, the application controls the risk of certificate number error in advance, reduces the subsequent transportation problems, and improves the accuracy required for the certificate number audit of air shipment and the audit efficiency of air shipment, reduces the transportation risk.

[0020] Especially, for the certificate number meeting the error allowance benchmark, the application further analyzes the alignment degree of the actual number corresponding to the certificate number and the compliance number, realizes the cross verification of digital information and physical information, improves the accuracy of the audit, in the air shipment, there are two cases of "input error" or "false reporting" in the certificate number presented in the cargo approval report, it is one-sided to judge only by the number characteristics in the received cargo approval report, calling the actual number of the cargo picture as the physical layer real basis, and comparing the marked compliance number through the alignment degree, quantifying the matching degree of the physical number and the standard number according to the alignment, effectively distinguishing "input error" and "false reporting", and then, the application avoids the misjudgment caused by only relying on the audit report number and ignoring the actual information of the goods, greatly improves the accuracy of the audit, especially can intercept the high-risk situation that the cargo approval report is false but the goods itself is not compliant, at the same time, greatly reduces the occupation of audit resources, makes the compliant goods pass the audit quickly, shortens the circulation period of the goods, the application controls the risk of certificate number error in advance, reduces the problem in the subsequent transportation link, and improves the accuracy required by the air shipment for the certificate number audit and the efficiency of the air shipment audit, reduces the transportation risk. BRIEF DESCRIPTION OF DRAWINGS

[0021] Figure 1 The figure shows the steps of the intelligent audit method for the cargo approval report of the embodiment of the application. Figure 2 The figure shows the logic judgment graph of the error abnormality degree category of the cargo approval report of the embodiment of the application. Figure 3 The figure shows the logic judgment graph of the evaluation of whether to meet the error allowance benchmark of the embodiment of the application. Figure 4 The figure shows the logic judgment graph of the processing and marking of the corresponding cargo approval report of the embodiment of the application. DETAILED DESCRIPTION

[0022] In order to make the purpose and advantages of the application more clear and obvious, the application will be further described below in combination with the embodiments; it should be understood that the specific embodiments described herein are only used to explain the application, and do not limit the protection scope of the application.

[0023] The preferred embodiments of the application will be described below with reference to the drawings. Those skilled in the art should understand that these embodiments are only used to explain the technical principles of the application, and are not intended to limit the protection scope of the application.

[0024] Please refer to Figure 1 The figure shows the steps of the intelligent audit method for the cargo approval report of the embodiment of the application. The intelligent audit method for the cargo approval report of the embodiment of the application comprises: Step S1, obtaining form data of the received cargo approval report of the receiving end, positioning the keywords in the form data, and identifying the number corresponding to the keywords; Step S2, calling the number matching feature corresponding to the cargo approval report applied by a single client in a single time to calculate the application error representation value of a single cargo approval report, and dividing the error abnormality degree category of the cargo approval report; Step S3, according to the error abnormality degree category, performing audit analysis on the cargo approval report, including, comparing the certificate number with the compliance number in the compliance number library, extracting the number error feature of the shipped goods, determining the approval deviation value of the certificate number, and evaluating whether it meets the error allowed benchmark, and processing the cargo approval report; Or, obtaining the proportion of the application error report in the cargo approval report applied by the client in a single time, determining whether to mark the client, and marking the application error report as a suspicious report; Wherein, the number matching feature includes number missing proportion and number coincidence degree, and the number error feature includes error character and compliance character space distance and minimum adjacent distance between error characters in the certificate number.

[0025] Specifically, the form data includes the number matching feature of the cargo approval report, the number error feature of the shipped goods, the proportion of the application error report in the cargo approval report applied by a single client in a single time, etc. The form data can be stored in a related database for data calling.

[0026] Specifically, the keyword refers to the "certificate number" in the cargo approval report. The form content of the cargo approval report is extracted by using OCR full image scanning, the coordinates of the keywords are recorded, the fixed range is expanded based on the keywords, the number area is intercepted, the number edge is extracted by fusing Canny edge operator and Roberts edge operator, so that the number edge is clearer, and the clear input is provided for OCR identification. The number of the application approval report is obtained by OCR identification, wherein the number is composed of characters including numbers and letters.

[0027] Specifically, the compliance number library is a database for storing compliance numbers by the receiving end, which will not be repeated here.

[0028] Specifically, the application analyzes the report applied by a single client in a single time, rather than mixing the cargo approval reports submitted by multiple clients or a single client multiple times, identifies the high-risk report submitted in a single time in real time, ensures to lock the risk at this time in the first time, and avoids risk lag due to waiting for "multiple data accumulation".

[0029] Specifically, the process of calculating the application error representation value of a single goods approval report includes: taking the ratio of the number missing proportion to the number missing proportion threshold value as the first application error feature; taking the ratio of the number coincidence degree threshold value to the number coincidence degree as the second application error feature; taking the sum of the first application error feature and the second application error feature as the application error representation value.

[0030] wherein, the certificate number is matched with the compliance number, the proportion of the number of missing characters of the certificate number relative to the total number of characters corresponding to the certificate number is analyzed as the number missing proportion, the certificate number is matched with the compliance number, the same characters are determined, and the ratio of the number of same characters to the total number of characters is taken as the number coincidence degree.

[0031] In this embodiment, the purpose of setting the number missing proportion threshold value and the number coincidence degree threshold value is to represent the case that the error abnormality degree of the number in the goods approval report is high. The relevant historical records of a plurality of completed audit goods approval reports of the same client are obtained, the number missing proportion historical data and the number coincidence degree historical data are called, the number missing proportion mean value and the number coincidence degree mean value are solved, and the corresponding are taken as the reference value under normal circumstances. Based on the purpose of setting the above two threshold values, the number missing proportion threshold value is determined as the product of the number missing proportion mean value and the missing deviation coefficient, and the number coincidence degree threshold value is determined as the product of the number coincidence degree mean value and the coincidence deviation coefficient. The missing deviation coefficient is selected within the interval [1.2, 1.25], and in the implementation, it is preferred to be 1.2. The coincidence deviation coefficient is selected within the interval [0.9, 0.95], and in the implementation, it is preferred to be 0.9.

[0032] Specifically, the present application considers the influence of the error of the certificate number on the transport compliance in the air shipment of goods, aggregates the "integrity defect" through the number missing proportion, directly reflects the proportion of the number of missing characters in the total number, can intuitively distinguish between "partial loss" and "large area loss", quantifies the possible damage degree of the core identification function of the number, for example, the number missing proportion is small (such as missing only 1 bit), which may be accidental omission, and the integrity impact is small; and the number missing proportion is too high, which may cause the number to lose the core identification function; focus on the "matching defect" through the number coincidence degree, reflect the matching proportion of the same characters in the certificate number and the compliance number, and judge whether the error is "local error" or "overall error", for example, the number coincidence degree between the number in the goods approval report and the number in the compliance number library is high, which indicates that the number has errors but the core information is not lost; and the coincidence degree is too low, which may be an overall error or a mixed number, the combination of the two can convert the fuzzy judgment of the "number error abnormality degree" into a quantifiable numerical feature, ensure the consistency and accuracy of the number error evaluation, and then, calculate the error characteristic value for subsequent division of the error abnormality degree category of the goods approval report, the present application controls the risk of certificate number error in advance, reduces the problems in the subsequent transportation link, improves the accuracy required for the audit of the certificate number of air shipment and the audit efficiency of air shipment, and reduces the transportation risk.

[0033] Specifically, please refer to Figure 2 As shown in the figure, it is a logic judgment diagram for dividing the error abnormality degree category of the goods approval report according to the embodiment of the present application, and the error abnormality degree category of the goods approval report is divided, including: If the error characteristic value of the goods approval report is greater than or equal to the error characteristic threshold value, the goods approval report is divided into a high error abnormality degree category; If the error characteristic value of the goods approval report is less than the error characteristic threshold value, the goods approval report is divided into a low error abnormality degree category.

[0034] The error characteristic threshold value is determined in advance, and the error characteristic value calculated under the condition that the number missing proportion is equal to the number missing proportion threshold value and the number coincidence degree threshold value is equal to the number coincidence degree is determined as the error characteristic threshold value.

[0035] Specifically, according to the error abnormality degree category, the goods approval report is subjected to audit analysis, including: If the goods approval report is in the low error abnormality degree category, the certificate number is compared with the compliance number in the compliance number library, the number error characteristics of the shipped goods are extracted, the approval deviation value of the certificate number is determined, and it is determined whether the error is within the error allowed benchmark, and the goods approval report is processed; If the cargo approval report is of the high error abnormality degree category, the proportion of the error report in the cargo approval report of the single application of the client is obtained, it is determined whether to mark the client, and the error report is marked as a suspicious report.

[0036] Specifically, the process of determining the approval deviation value of the certificate number includes: calculating a spatial deviation value of the spatial distance of the error character and the compliance character and a spatial distance threshold value; calculating an adjacency deviation value of the minimum adjacency distance between the error characters in the certificate number and a minimum adjacency distance threshold value; taking the sum of the spatial deviation value and the adjacency deviation value as the approval deviation value.

[0037] In this embodiment, the spatial deviation value and the adjacency deviation value are calculated in the following manner: calculating a spatial difference absolute value of the spatial distance of the error character and the compliance character and a spatial distance threshold value, taking the ratio of the spatial difference absolute value and the spatial distance threshold value as the spatial deviation value; correspondingly, calculating an adjacency difference absolute value of the minimum adjacency distance between the error characters in the certificate number and a minimum adjacency distance threshold value; taking the ratio of the adjacency difference absolute value and the minimum adjacency distance threshold value as the adjacency deviation value.

[0038] Wherein, the purpose of setting the spatial distance threshold value and the minimum adjacency distance threshold value is to represent the case that the severity of the number error is higher and the error presented by the number has a higher possibility of potential risk. Obtain the relevant historical records of several completed cargo approval reports of the same client, call the spatial distance historical data of the error character and the compliance character and the minimum adjacency distance historical data between the error characters in the certificate number, solve the spatial distance mean and the minimum adjacency distance mean, and correspondingly as the reference value under normal circumstances. Based on the purpose of setting the above two threshold values, the spatial distance threshold value is determined as the product of the spatial distance mean and the spatial deviation coefficient, and the minimum adjacency distance threshold value is determined as the product of the minimum adjacency distance mean and the adjacency deviation coefficient, wherein the spatial deviation coefficient is selected within the interval [1.4, 1.6], and the implementation is preferred to be 1.4, and the adjacency deviation coefficient is selected within the interval [1.2, 1.3], and the implementation is preferred to be 1.2.

[0039] In this embodiment, the character in the number that does not correspond to the character in the compliance number is taken as the error character, which will not be repeated here.

[0040] Specifically, for the low error abnormality degree category, the cargo approval report is finely reviewed to avoid labeling the correct numbered cargo report as a suspicious report, the present application quantifies the contingency of the input error by the spatial distance between the error character and its corresponding compliance character, avoids excessive interception of "non-subjective errors", for example, if the error character and the compliance character are close in position distance in the keyboard space, it means that the error is more likely to be an accidental error of "finger mis-touching the keys close in distance"; if the position distance is far, it may be a subjective error of "memory error or intentional error"; the correlation of the error is judged by analyzing the minimum adjacent distance between the error characters in the certificate number, to distinguish whether the formed error is a "concentrated error" (may be a systematic problem) or a "dispersed touch" (accidental error), for example, if the error characters are close in distance or there is only a single error character, it belongs to the case of "local dispersed touch" (such as temporary misreading during input), the risk is low; and if the error characters are far apart in distance, it means that the error characters are dispersed in the number, which may be "unfamiliar with the numbering rules" (such as misremembering multiple characters) or "template filling error", which belongs to potential systematic error, the risk is high, therefore, the present application determines the approval deviation value of the certificate number by considering the above-mentioned situations, quantifies the severity of the error itself, indirectly evaluates the potential risk of the error, provides data support for subsequent evaluation of whether it meets the error allowance benchmark, the present application controls the risk of certificate number error in advance, reduces the problems in the subsequent transportation link, improves the accuracy required for the certificate number review of air transportation and the review efficiency of air transportation, and reduces the transportation risk.

[0041] Specifically, please refer to Figure 3 The present application is a logical decision diagram for evaluating whether the error allowance benchmark is met, and the evaluation of whether the error allowance benchmark is met comprises: If the approval deviation value of the certificate number is greater than or equal to the approval deviation threshold, the certificate number is evaluated as not meeting the error allowance benchmark; If the approval deviation value of the certificate number is less than the approval deviation threshold, the certificate number is evaluated as meeting the error allowance benchmark.

[0042] The approval deviation threshold is predetermined, the spatial deviation value when the spatial distance between the error character and the compliance character is equal to the spatial distance threshold, and the adjacent deviation value when the minimum adjacent distance between the error characters in the certificate number is equal to the minimum adjacent distance threshold, are calculated, and the sum of the spatial deviation value and the adjacent deviation value is taken as the approval deviation threshold.

[0043] Specifically, the cargo approval report is processed, comprising: If the certificate number meets the error tolerance benchmark, mark the corresponding compliance number, call the certificate number corresponding to the picture of the shipped goods, analyze the actual number of the shipped goods, determine the number alignment degree of the actual number and the compliance number, and mark the corresponding goods approval report for processing; If the certificate number does not meet the error tolerance benchmark, mark the goods approval report corresponding to the certificate number as a suspicious report.

[0044] In this embodiment, by taking the number of other characters between error characters + 1 as the adjacent spacing of error characters, the minimum adjacent distance between error characters in the certificate number is determined; Wherein, for the case of 0 error characters and only 1 error character, since there is no "error characters with a certain distance", in the implementation, when the error character is 0, the minimum adjacent distance is determined as 0, and when there is only one error character, the minimum adjacent distance is determined as 1, which will not be repeated here.

[0045] Specifically, the process of determining the number alignment degree of the actual number and the compliance number includes: The character corresponding number ratio of the actual number and the compliance number is taken as the number alignment degree.

[0046] It can be understood that when the correctness of the actual number and the compliance number is audited, it is necessary to check whether each character in the two numbers is one-to-one corresponding and same, for example, according to the order from left to right, if the actual number and the compliance number are completely matched, then the characters of the actual number and the single characters of any corresponding bit of the compliance number from left to right are same, based on this, the number of characters corresponding one-to-one between the characters of the actual number and the characters of the compliance number is determined, the ratio of the character number and the total number of characters is taken as the number alignment degree, and the number alignment degree threshold is set to 85%.

[0047] Specifically, please refer to Figure 4 The above is the logical decision diagram of the present application for marking the corresponding goods approval report for processing, marking the corresponding goods approval report for processing, including: If the number alignment degree of the actual number and the compliance number is greater than or equal to the number alignment degree threshold, it is determined that the corresponding goods approval report can be modified; If the number alignment degree of the actual number and the compliance number is less than the number alignment degree threshold, the corresponding goods approval report is marked as a suspicious report.

[0048] Specifically, for the certificate number meeting the error allowance benchmark, the application further analyzes the alignment degree of the actual number corresponding to the certificate number and the compliance number, realizes cross verification of digital information (approval report number) and physical information (actual number of shipped goods), improves the accuracy of the audit, and in air shipment, there are two cases of "input error" (such as accidental touch) or "false reporting" (such as intentionally filling in an incorrect number) in the certificate number presented in the goods approval report. It is one-sided to judge only by the number characteristics in the received goods approval report. The actual number extracted from the picture of the shipped goods is called as a physical level of true basis, and the marked compliance number is compared by alignment degree. The matching degree of physical number and standard number according to sequential alignment is quantified, which effectively distinguishes "input error" and "false reporting". If the certificate number deviates from the compliance number due to input error, but the alignment degree of the actual number of the shipped goods picture and the compliance number is high, it means that "the goods approval report is filled in wrong but the goods itself is compliant" (such as accidental touch). The corresponding goods approval report can be modified to avoid the delay of the shipped goods due to minor errors. If the alignment degree of the certificate number and the actual number is low, it means that "the goods may not match the declaration" (such as wrong goods, false declaration), the goods approval report needs to be marked, and the goods approval report can be reviewed in the future (such as checking whether the goods are wrong or whether there is false declaration), reducing the range of manual intervention. Further, the application avoids the misjudgment caused by relying only on the audit report number and ignoring the actual information of the goods through three comparisons, greatly improves the accuracy of the audit, especially can intercept the high-risk situation of false goods approval report but the goods itself is not compliant, at the same time, greatly reduces the occupation of audit resources, so that the compliant goods can pass the audit quickly, shortens the goods circulation period, the application controls the risk of certificate number error in advance, reduces the problems in the subsequent transportation link, and improves the accuracy required by air shipment for certificate number audit and the efficiency of air shipment audit, reduces the transportation risk.

[0049] Specifically, determining whether to mark the client includes: If the proportion of error reports is greater than or equal to the proportion threshold, the client is marked; Wherein, the goods approval report with a request error representation value greater than or equal to a request error representation threshold is determined as the error report.

[0050] In the embodiment, the purpose of setting the proportion threshold is to represent the case that the number of error report in the single client's single application for the goods approval report accounts for a large proportion, the corresponding number anomaly is more serious, and the client has a large possibility of abnormality. The relevant historical records of the client's several completed goods approval reports are obtained, the proportion of error report history data is called, the proportion mean is solved, and is used as the reference value in the normal case. The proportion threshold is determined as the product of the proportion mean and the proportion deviation coefficient, wherein the proportion deviation coefficient is selected in the interval [1.2, 1.3], and 1.2 is preferred in the implementation.

[0051] If the intelligent auditing method for the goods approval report of the present application is realized in the form of a software function 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 application or the part that essentially contributes to the prior art or the whole or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device) to execute all or part of the steps of the method described in the various embodiments of the present application. The aforementioned storage medium includes a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various program code storage media.

[0052] So far, the technical solution of the present application has been described in combination with the preferred embodiments shown in the drawings, but those skilled in the art can easily understand that the protection scope of the present application is obviously not limited to these specific embodiments. Those skilled in the art can make equivalent changes or replacements to the related technical features without departing from the principles of the present application, and the technical solutions after the changes or replacements will fall within the protection scope of the present application.

Claims

1. An intelligent auditing method for cargo vetting report, characterized in that, The method comprises the following steps: Obtaining form data of a cargo approval report received by a receiving end, locating a keyword in the form data, and identifying a number corresponding to the keyword; Calling a number matching feature corresponding to a cargo approval report submitted by a single client in a single time to calculate a submission error representation value of a single cargo approval report, and dividing the error abnormality degree category of the cargo approval report; According to the error abnormality degree category, the cargo approval report is analyzed, including, Comparing the certificate number with the compliance number in the compliance number library, extracting the number error feature of the transported cargo, determining the approval deviation value of the certificate number, and evaluating whether it meets the error allowance benchmark to process the cargo approval report; Or, obtaining the proportion of the submission error report in the cargo approval report submitted by the client in a single time, determining whether to mark the client, and marking the submission error report as a suspicious report; Wherein, the number matching feature includes the number missing proportion and the number coincidence degree, and the number error feature includes the spatial distance of error characters and compliance characters and the minimum adjacent distance between error characters in the certificate number. 2.The intelligent auditing method for cargo approval report according to claim 1, wherein, The process of calculating the submission error representation value of a single cargo approval report includes: Taking the ratio of the number missing proportion to the number missing proportion threshold value as the first submission error feature; Taking the ratio of the number coincidence degree threshold value to the number coincidence degree as the second submission error feature; Taking the sum of the first submission error feature and the second submission error feature as the submission error representation value. 3.The intelligent auditing method for cargo vetting report according to claim 2, wherein, Dividing the error abnormality degree category of the cargo approval report includes: If the submission error representation value of the cargo approval report is greater than or equal to the submission error representation threshold value, the cargo approval report is divided into a high error abnormality degree category; If the submission error representation value of the cargo approval report is less than the submission error representation threshold value, the cargo approval report is divided into a low error abnormality degree category.

4. The intelligent auditing method for cargo vetting report according to claim 3, wherein, According to the error abnormality degree category, the cargo approval report is analyzed, including: If the cargo approval report is in the low error abnormality degree category, the certificate number is compared with the compliance number in the compliance number library, the number error feature of the transported cargo is extracted, the approval deviation value of the certificate number is determined, and it is evaluated whether it meets the error allowance benchmark to process the cargo approval report; If the cargo approval report is in the high error abnormality degree category, the proportion of the submission error report in the cargo approval report submitted by the client in a single time is obtained, it is determined whether to mark the client, and the submission error report is marked as a suspicious report.

5. The intelligent auditing method for cargo vetting report as claimed in claim 1 wherein, The process of determining the approval deviation value of the certificate number includes: Calculating the spatial deviation value of the spatial distance of error characters and compliance characters and the spatial distance threshold value; Calculating the adjacent deviation value of the minimum adjacent distance between error characters in the certificate number and the minimum adjacent distance threshold value; Taking the sum of the spatial deviation value and the adjacent deviation value as the approval deviation value.

6. The intelligent auditing method for cargo vetting report as claimed in claim 5 wherein, Evaluating whether it meets the error allowance benchmark includes: If the approval deviation value of the certificate number is greater than or equal to the approval deviation threshold value, the certificate number is evaluated as not meeting the error allowance benchmark; If the approval deviation value of the certificate number is less than the approval deviation threshold value, the certificate number is evaluated as meeting the error allowance benchmark.

7. The intelligent auditing method for cargo vetting report according to claim 6, wherein, The cargo approval report is processed, including: If the certificate number meets the error allowance benchmark, the corresponding compliance number is marked, the carrier cargo picture corresponding to the certificate number is called, the actual number of the carrier cargo is analyzed, the number alignment degree of the actual number and the compliance number is determined, and the corresponding cargo approval report is processed and marked. If the certificate number does not meet the error allowance benchmark, the cargo approval report corresponding to the certificate number is marked as a suspicious report. 8.The intelligent auditing method for cargo approval report according to claim 7, wherein, The process of determining the number alignment degree of the actual number and the compliance number includes: The character corresponding number ratio of the actual number and the compliance number is taken as the number alignment degree. 9.The intelligent auditing method for cargo vetting report according to claim 8, wherein, The corresponding cargo approval report is processed and marked, including: If the number alignment degree of the actual number and the compliance number is greater than or equal to the number alignment degree threshold value, it is determined that the corresponding cargo approval report can be modified; If the number alignment degree of the actual number and the compliance number is less than the number alignment degree threshold value, the corresponding cargo approval report is marked as a suspicious report.

10. The intelligent auditing method for cargo vetting report as claimed in claim 1 wherein, The determination of whether to mark the client includes: If the proportion of error reports is greater than or equal to the proportion threshold value, the client is marked; Wherein, the cargo approval report with the error report representation value greater than or equal to the error report representation threshold value is determined as the error report.

Images