Service data management method for cross-border e-commerce

Abstract

The invention discloses a cross-border e-commerce service data management method, and relates to the technical field of service data management, and the method comprises the steps: obtaining commodity service data from a plurality of overseas e-commerce platforms, and caching the commodity service data to an edge data pool according to the attributes of a data source; based on source attributes, updating time and historical accuracy of the data sources and supervision rating of judicial jurisdictions, comprehensive credible weight values of the data sources are calculated, and main data channels are screened out according to the weight values; inputting commodity service data in the main data channel into a multi-language semantic encoder, generating semantic vectors, calculating cross-language semantic alignment, and comparing logic consistency of key structured fields; generating a consistency score according to a comparison result of the semantic alignment degree and the structured field, when the score is lower than a preset threshold value, marking as abnormal and entering a rechecking process, and otherwise, entering a compliance review stage; and calling the regional compliance rule base to carry out regulation matching on the data passing the consistency verification.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of service data management, in particular to a service data management method for cross-border e-commerce. BACKGROUND

[0002] Service data management technology refers to a series of technologies and methods used to collect, process, analyze and optimize data related to goods and services in the e-commerce environment, especially in the field of cross-border e-commerce. The technology aims to improve data quality, ensure data security, promote data sharing, and support more informed business decisions based on data analysis. Therefore, how to use advanced technical means to improve the intelligent level and security of service data management has become one of the problems to be solved.

[0003] In the field of service data management technology, since the goods and service data usually need to be obtained from e-commerce platforms distributed in different regions around the world, it will result in long cross-border transmission time and high bandwidth cost, and there are translation deviations and information omissions in different language versions of goods information, which affects user experience and purchase decision. SUMMARY

[0004] In view of the above existing problems, the present application is proposed.

[0005] Therefore, the present application provides a service data management method for cross-border e-commerce to solve the problem that the goods and service data usually need to be obtained from e-commerce platforms distributed in different regions around the world, which results in long cross-border transmission time and high bandwidth cost, and there are translation deviations and information omissions in different language versions of goods information, which affects user experience and purchase decision.

[0006] To solve the above technical problems, the present application provides the following technical solutions: In a first aspect, the present application provides a service data management method for cross-border e-commerce, which comprises: Obtaining goods and service data from multiple overseas e-commerce platforms and caching them to an edge data pool according to data source attributes; Calculating the comprehensive trust weight value of each data source based on the source attribute, update time, historical accuracy rate and supervision rating of the jurisdiction of the data source, and selecting the main data channel according to the weight value; Inputting the goods and service data in the main data channel into a multi-language semantic encoder to generate semantic vectors and calculate cross-language semantic alignment, and comparing the logical consistency of key structured fields; Generating a consistency score according to the semantic alignment and structured field comparison results, marking as abnormal and entering the review process when the score is lower than the preset threshold, otherwise entering the compliance review stage; The compliance rule library is called to perform regulation matching on the data passing the consistency check, and outputs a compliance score; data with a score lower than a threshold is prohibited from synchronization and a risk log is recorded; For data passing the compliance review, a scheduling priority is generated in combination with the trusted weight, consistency score, transmission delay and target market active period; Cross-border data synchronization is performed according to the scheduling priority, and the trusted weight and consistency evaluation model parameters are dynamically optimized using user behavior feedback data to form a closed-loop updating mechanism.

[0007] As a preferred scheme of the service data management method for cross-border e-commerce, the service data of the goods from the plurality of overseas e-commerce platforms is obtained and cached to an edge data pool according to data source attributes, specifically including: Periodically capturing the service data of the goods from the e-commerce platform interfaces distributed in different countries and regions, the service data of the goods including the name of the goods, the price, the delivery policy, the return and exchange instructions and the language description text of the target market; Attaching metadata tags to the captured data, the metadata tags including the platform identification of the data source, the data collection timestamp, the country code of the target market, the original language type and the IP geographical location of the data provider; Temporarily storing the service data of the goods with the metadata tags in a cache queue deployed in a regional edge node to form an edge data pool, the edge node being located in the geographical region or the adjacent region of the data source, for reducing the cross-border transmission delay.

[0008] As a preferred scheme of the service data management method for cross-border e-commerce, the comprehensive trusted weight value of each data source is calculated based on the source attributes of the data source, the update time, the historical accuracy rate and the supervision rating of the jurisdiction area, and the main data channel is selected according to the weight value, specifically including: Extracting the platform identification, the collection timestamp and the IP geographical location in the metadata tags, and combining the pre-established data source historical accuracy rate database to obtain the data deviation frequency of the data source in the past plurality of processing periods, and calculating the historical accuracy rate; Querying the preset supervision rating table of the jurisdiction area to obtain the supervision strength score of the country or region where the data source is located; According to the difference between the data collection timestamp and the current time, a timeliness decay factor is calculated; The historical accuracy rate, the supervision strength score, the timeliness decay factor and the consistency state of the IP geographical location and the platform registration place are taken as input variables; The comprehensive trusted weight value is calculated by weighted summation , and the expression is: ; wherein, is a comprehensive trust weight value, is a trust component, is a timeliness decay factor, is a regulatory strength score, is a geographic trust state value, , , , is a preset weight coefficient; sort all data sources according to value, select the data stream corresponding to the data source greater than the first threshold value as the main data channel, and the rest of the data enters the verification queue.

[0009] As a preferred scheme of the service data management method of the cross-border e-commerce, wherein: the commodity service data in the main data channel is input into a multi-language semantic encoder to generate a semantic vector and calculate a cross-language semantic alignment degree, and the logical consistency of the key structured fields is compared, specifically including: extracting the description texts of the same commodity in two different language versions in the main data channel, inputting the pre-trained multi-language semantic encoding model respectively, and obtaining the corresponding high-dimensional semantic vectors; calculating the cosine similarity between the two semantic vectors as the cross-language semantic alignment degree index; extracting the key structured fields corresponding to the texts, including the price value, the delivery range and the return and exchange period; judging whether the relative difference of the price value is less than a preset tolerance coefficient, whether the delivery range covers the same area, and whether the return and exchange period is within the allowed deviation range; if the fields all meet the consistency condition, it is determined that the structured information is consistent; otherwise, it is determined that it is inconsistent; the semantic alignment degree and the structured information consistency state are taken as the input basis of the subsequent consistency score.

[0010] As a preferred scheme of the service data management method of the cross-border e-commerce, wherein: the consistency score is generated according to the semantic alignment degree and the comparison result of the structured fields, and when the score is lower than a preset threshold, it is marked as abnormal and enters the review process, otherwise it enters the compliance review stage, specifically including: fusion calculation is performed on the obtained semantic alignment degree and structured information consistency state; a linear weighting method is used to generate a consistency score , the expression is: ; wherein, is a consistency score, ​a semantic alignment degree, a structured information consistency flag, a semantic weight coefficient; the data is determined to be consistent across languages and enters a compliance review stage if the semantic weight coefficient is not lower than the second threshold value; the data is marked as a to-be-reviewed state if the semantic weight coefficient is lower than the second threshold value, a difference report containing a difference field and original text is generated, and the data is pushed to an artificial review process.

[0011] As a preferred solution of the cross-border e-commerce service data management method, the data that passes the consistency check is subjected to regulation matching by calling a regional compliance rule library, and a compliance score is outputted, and data with a score lower than a threshold value is prohibited from synchronization and a risk log is recorded, specifically including: According to the target market country code, the jurisdiction area where the data is to be published is determined; The compliance rule set corresponding to the jurisdiction area is called from the preset regional compliance rule library, and the compliance rule set includes data privacy, advertising, product access, and consumer rights protection clauses; Whether the commodity service data meets each requirement in the compliance rule set is compared item by item, and the number of items that meet the requirement is counted; The ratio of the number of items that meet the requirement to the total number of rules is taken as the compliance score; The compliance score is compared with a third threshold value, and if the compliance score is lower than the third threshold value, the data is prohibited from participating in cross-border synchronization operation, and the rule number and field that violate the regulation are recorded in the risk traceability log; If the compliance score is not lower than the third threshold value, the data is allowed to enter a dispatch queue, and the compliance score of the data is taken as one of the input parameters for priority calculation.

[0012] As a preferred solution of the cross-border e-commerce service data management method, the data that passes the compliance review is combined with its trust weight, consistency score, transmission delay, and target market active period to generate a dispatch priority, specifically including: The predicted transmission delay of the data from an edge node to a target market front-end system is obtained; The state identifier of whether the target market is currently in a user access peak period is obtained; The comprehensive trust weight value, consistency score, and compliance score are normalized to obtain a standardized value; A business conversion potential coefficient is introduced and determined according to the commodity category and operation state; A preliminary data value index is calculated by using a weighted summation method , and the expression is:​​​ ; wherein is a preliminary data value index, , , respectively are normalized trust weight, consistency score and compliance score, is a business conversion potential coefficient, , , , is a weight coefficient; divide the preliminary data value index by the transmission delay, and combine the user active period state to carry out gain adjustment, to obtain a scheduling priority; The higher the scheduling priority is, the more preferentially the data is executed cross-border synchronization.

[0013] As a preferred scheme of the cross-border e-commerce service data management method, wherein: the cross-border data synchronization is executed according to the scheduling priority, and the trust weight and the consistency evaluation model parameters are dynamically optimized by using the user behavior feedback data to form a closed-loop updating mechanism, specifically including: all data passing through the compliance review are sorted according to the final scheduling priority from high to low to form a scheduling queue; data is taken out from the scheduling queue in turn and cross-border transmission is executed, and is pushed to the front-end display system or the operation database of the corresponding country region; After the data is put online, the interactive behavior data of the target market users is collected, including the number of clicks, the page stay duration, the number of order conversion and the number of complaints; The user feedback score is calculated according to the interactive behavior data, reflecting the actual performance of the data quality; The user feedback score is taken as a real label, and regression analysis is carried out with the comprehensive trust weight value and the consistency score to adjust the weight parameters in the trust model and the consistency evaluation model; The updated model parameters are written into the system configuration for the next round of data processing period, to realize dynamic iterative optimization of the model parameters.

[0014] In the second aspect, the present application provides a computer device, comprising a memory and a processor, and the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of the cross-border e-commerce service data management method of the first aspect of the present application is realized.

[0015] In the third aspect, the present application provides a computer readable storage medium, which stores a computer program, wherein: when the computer program is executed by the processor, any step of the cross-border e-commerce service data management method of the first aspect of the present application is realized.

[0016] The application has the beneficial effects that: by constructing a multi-dimensional dynamic weight credibility evaluation model, combining cross-language semantic alignment and structured field consistency checking mechanism, and fusing priority scheduling strategy under regional compliance constraints, intelligent, accurate and compliant management of cross-border e-commerce service data is realized, effectively improving the screening accuracy of heterogeneous data sources, the consistency guarantee capability of multi-language commodity information, and the legality and timeliness of cross-border data synchronization, and through user behavior feedback driving model parameter dynamic optimization, a closed-loop management mechanism is formed, significantly enhancing the system adaptive ability and data governance level, and achieving significant technical effects in ensuring user experience and avoiding operation risks. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0018] Figure 1 Flowchart of the service data management method of cross-border e-commerce in embodiment 1. DETAILED DESCRIPTION

[0019] In order to make the above-mentioned purposes, features and advantages of the present application more apparent and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the drawings of the specification.

[0020] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application, but the present application can also be implemented in other ways different from those described herein, and those skilled in the art can make similar generalizations without departing from the connotation of the present application, therefore the present application is not limited to the specific embodiments disclosed below.

[0021] Secondly, the "one embodiment" or "embodiment" referred to herein means that the specific features, structures or characteristics can be included in at least one implementation of the present application. "In one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an independent or alternative embodiment that excludes other embodiments.

[0022] Embodiment, refer to Figure 1 For the embodiments of the present application, the embodiment provides a service data management method of cross-border e-commerce, comprising the following steps: S1, obtaining commodity service data from a plurality of overseas e-commerce platforms, and caching them to an edge data pool according to data source attributes; Further, periodically grab commodity service data from e-commerce platform interfaces distributed in different countries and regions, including commodity name, price, delivery policy, return and exchange instructions, and target market language description text; Attach metadata tags to the grabbed data, including data source platform identification, data collection timestamp, target market country code, original language type, and data provider IP geographic location; Temporarily store commodity service data with metadata tags in a cache queue deployed in a regional edge node to form an edge data pool. The edge node is located in the geographic region or adjacent region of the data source, used to reduce cross-border transmission delay; It should be noted that by establishing a data caching mechanism in the edge node, preliminary data aggregation and tagging can be completed at the network edge close to the data source, effectively reducing the real-time access pressure of the central server, and reducing network delay and bandwidth consumption caused by long-distance cross-border transmission., provide stable and low-latency data input basis for efficient screening and processing of subsequent multi-source data.

[0023] S2, based on the source attribute, update time, historical accuracy rate and jurisdictional supervision rating of the data source, calculate the comprehensive trust weight value of each data source, and select the main data channel according to the weight value; Further, extract the platform identification, collection timestamp, and IP geographic location from the metadata tags, and combine the pre-established data source historical accuracy rate database to obtain the data deviation frequency of the data source in the past multiple processing periods, and calculate the historical accuracy rate; Query the pre-set jurisdictional supervision rating table to obtain the supervision intensity score of the country or region where the data source is located; According to the difference between the data collection timestamp and the current time, calculate the timeliness decay factor; Take the historical accuracy rate, supervision intensity score, timeliness decay factor, and consistency status of the IP geographic location and platform registration location as input variables; Calculate the comprehensive trust weight value using weighted summation, the expression is: ; Wherein, is the comprehensive trust weight value, is the trust component, is the timeliness decay factor, is the supervision intensity score, is the geographic trust state value, , , , is the preset weight coefficient; According to all data sources Value sorting, selection Data streams corresponding to data sources greater than the first threshold value are taken as the main data channel, and the rest of the data enters the verification queue; It should be noted that the construction of the comprehensive trust weight value combines multiple dimensions such as the historical performance of the data source, the regulatory environment, timeliness, and geographical trustworthiness, avoiding the misjudgment risk brought by relying on a single indicator for data screening, improving the scientificity and robustness of the selection of the main data channel, and ensuring that the data entering the core processing process has high original quality and trustworthiness.

[0024] S3, input the commodity service data in the main data channel into the multilingual semantic encoder to generate semantic vectors and calculate the cross-language semantic alignment degree, and compare the logical consistency of the key structured fields; Further, the description texts of the same commodity in two different language versions in the main data channel are extracted and input into the pre-trained multilingual semantic encoding model respectively to obtain corresponding high-dimensional semantic vectors; The cosine similarity between the two semantic vectors is calculated as the cross-language semantic alignment degree index; The key structured fields corresponding to the text are extracted, including price value, delivery range, and return and exchange period; Determine whether the relative difference of the price value is less than the preset tolerance coefficient, whether the delivery range covers the same area, and whether the return and exchange period is within the allowed deviation range; If the fields all meet the consistency condition, it is determined that the structured information is consistent; otherwise, it is determined to be inconsistent; The semantic alignment degree and the structured information consistency state are used as the input basis for subsequent consistency scoring; It should be noted that the multilingual semantic encoding model is used to vectorize the commodity description text, and the logical comparison of the structured fields is combined to realize double consistency verification of the commodity information at the semantic level and the data level, which can effectively identify the cross-language description inconsistency problem caused by translation deviation, information omission or intentional misleading, and improve the authenticity and comparability of cross-border commodity information.

[0025] S4, generate a consistency score according to the semantic alignment degree and the structured field comparison result, and mark it as abnormal and enter the review process when the score is lower than the preset threshold, otherwise enter the compliance review stage; Further, the obtained semantic alignment degree and structured information consistency state are fused and calculated; A linear weighting method is used to generate a consistency score, and the expression is: ; Wherein, is the consistency score, is the semantic alignment degree, For the consistency of structured information, For the semantic weight coefficient; Will Compared with the second threshold, if Not less than the threshold, it is determined that the commodity service data has cross-language consistency, and enters the compliance review stage; If Lower than the threshold, it is marked as a pending review state, a difference report containing the difference field and the original text is generated, and it is pushed to the artificial audit process; It should be pointed out that the consistency scoring mechanism realizes the quantitative evaluation of the consistency of the complete description of the commodity service data by weighted fusion of semantic similarity and structured field matching results. The score as a precondition for entering the compliance review can identify and intercept data with potential information conflicts in the early stage, prevent incorrect or ambiguous content from entering the subsequent process, and ensure the reliability of the data processing chain.

[0026] S5, calling the regional compliance rule library to perform regulation matching on the data passing the consistency check, outputting a compliance score, and prohibiting synchronization of data with a score lower than the threshold and recording risk logs; Further, according to the target market country code, the jurisdiction where the data is intended to be published is determined; The corresponding compliance rule set of the jurisdiction is called from the pre-set regional compliance rule library, and the compliance rule set includes data privacy, advertising, product access and consumer protection provisions; Compare item by item whether the commodity service data meets the requirements in each item in the compliance rule set, and count the number of items that meet the requirements; The ratio of the number of items that meet the requirements to the total number of rules is taken as the compliance score; Compare the compliance score with the third threshold, if lower than the threshold, prohibit the data from participating in cross-border synchronization operation, and record the violation rule number and field in the risk traceability log; If not less than the threshold, allow the data to enter the dispatch queue, and take its compliance score as one of the input parameters for dispatch priority calculation; It should be pointed out that the introduction of the regional compliance rule library enables the system to accurately match different countries and regions according to laws and regulations, realize automatic verification of local compliance, avoid the risk of operation violation due to cultural differences or lagging legal updates, and significantly improve the legality and compliance response efficiency of cross-border e-commerce data publishing.

[0027] S6, for the data passing the compliance review, generate a dispatch priority combining the trusted weight, consistency score, transmission delay and target market active period; Further, the predicted transmission delay of the data from the edge node to the target market front-end system is obtained; obtaining a state identifier indicating whether the target market is currently in a user access peak period; normalizing the comprehensive trust weight value, the consistency score and the compliance score to obtain standardized values; introducing a business conversion potential coefficient, which is determined according to the commodity category and the operation state; calculating a preliminary data value index, using a weighted summation method, with the expression being: wherein is the preliminary data value index, , , are the normalized trust weight, the consistency score and the compliance score respectively, is the business conversion potential coefficient, , , , are weight coefficients; dividing the preliminary data value index by the transmission delay, and combining the user active period state to perform gain adjustment to obtain a scheduling priority; The higher the scheduling priority is, the more preferentially the data is executed for cross-border synchronization. It should be noted that the generation of the scheduling priority not only considers data quality factors, but also combines transmission efficiency and user active period and other operation scene characteristics, realizing the change from passive synchronization to intelligent optimization and pushing, ensuring that high-value and high-timeliness commodity information can reach target users in the best time window, maximizing the business conversion effect of data synchronization.

[0028] S7, performing cross-border data synchronization according to the scheduling priority, and dynamically optimizing the trust weight and the consistency evaluation model parameters using user behavior feedback data to form a closed-loop updating mechanism; Further, all data passing the compliance review are sorted in descending order of the final scheduling priority to form a scheduling queue; The data are sequentially taken out from the scheduling queue and executed for cross-border transmission to be pushed to the front-end display system or the operation database in the corresponding country region; After the data are put online, interactive behavior data of target market users are collected, including the number of clicks, the page stay duration, the number of order conversion and the number of complaints; According to the interactive behavior data, a user feedback score is calculated to reflect the actual performance of the data quality; The user feedback score is used as a real label to perform regression analysis with the comprehensive trust weight value and the consistency score to adjust the weight parameters in the trust model and the consistency evaluation model; ​Write the updated model parameters into the system configuration for the next round of data processing cycle, realize the dynamic iteration optimization of model parameters; It should be noted that the user behavior feedback and the closed-loop optimization mechanism of the model parameters enable the system to have the ability of continuous learning and self-evolution, can dynamically adjust the credible evaluation and consistency judgment strategy according to the actual market response, gradually improve the adaptability and prediction accuracy of the model to the real business scene, and form a long-term sustainable optimization data governance system.

[0029] The embodiment also provides a computer device suitable for the service data management method of cross-border e-commerce, which includes a memory and a processor.

[0030] The computer device can be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with external terminals in a wired or wireless manner. The wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the shell of the computer device, or an external keyboard, touchpad or mouse, etc.

[0031] The embodiment also provides a storage medium on which a computer program is stored, the program being executed by a processor to implement the service data management method for cross-border e-commerce proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic memory, a flash memory, a magnetic disk, or an optical disk.

[0032] To sum up, the application realizes intelligent, accurate and compliant management of cross-border e-commerce service data by constructing a multi-dimensional dynamic weight credibility evaluation model, combining cross-language semantic alignment and structured field consistency verification mechanism, and fusing priority scheduling strategy under regional compliance constraints, effectively improves the screening accuracy of heterogeneous data sources, the consistency guarantee capability of multi-language commodity information, and the legality and timeliness of cross-border data synchronization, and simultaneously forms a closed-loop management mechanism through dynamic optimization of model parameters driven by user behavior feedback, significantly enhances the system adaptive ability and data governance level, and achieves significant technical effects in terms of guaranteeing user experience and avoiding operation risks.

[0033] It should be noted that the above embodiments are only used to illustrate the technical solutions of the application rather than limit the application. Although the application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the application can be modified or replaced equivalently without departing from the spirit and scope of the technical solutions of the application, and all of them should be covered in the scope of the claims of the application.

Claims

1. A service data management method for cross-border e-commerce, characterized in that: The method comprises the following steps: Obtain commodity service data from multiple overseas e-commerce platforms and cache them to the edge data pool according to the data source attributes; Based on the source attributes, update time, historical accuracy rate and supervision rating of the jurisdiction area of the data source, calculate the comprehensive trust weight value of each data source, and filter out the main data channel according to the weight value; Input the commodity service data in the main data channel into the multi-language semantic encoder to generate semantic vectors and calculate the cross-language semantic alignment degree, and compare the logical consistency of the key structured fields; According to the semantic alignment degree and the comparison result of the structured field, generate a consistency score, and when the score is lower than the preset threshold, mark it as abnormal and enter the review process, otherwise enter the compliance review stage; Call the regional compliance rule library to match the data that passes the consistency check with the regulations, output the compliance score, and data with a score lower than the threshold is prohibited from synchronization and risk logs are recorded; For data that passes the compliance review, combine the trust weight, consistency score, transmission delay and target market active period to generate a scheduling priority; According to the scheduling priority, perform cross-border data synchronization, and dynamically optimize the trust weight and consistency evaluation model parameters using user behavior feedback data to form a closed-loop update mechanism. 2.The service data management method of cross-border e-commerce of claim 1, wherein: The method comprises the following steps: Periodically grab commodity service data from e-commerce platform interfaces distributed in different countries and regions, including commodity name, price, delivery policy, return and exchange instructions, and target market language description text; Add metadata tags to the grabbed data, including data source platform identification, data collection timestamp, target market country code, original language type, and data provider IP geographic location; Temporarily store the commodity service data with metadata tags in the cache queue deployed in the regional edge node to form an edge data pool, and the edge node is located in the geographic region or adjacent region of the data source to reduce cross-border transmission delay. 3.The service data management method of cross-border e-commerce of claim 2, wherein: The method comprises the following steps: Extract the platform identification, collection timestamp and IP geographic location from the metadata tags, and combine the pre-established data source historical accuracy rate database to obtain the data deviation frequency of the data source in the past multiple processing periods to calculate the historical accuracy rate; Query the preset jurisdiction area supervision rating table to obtain the supervision intensity score of the country or region where the data source is located; Calculate the timeliness decay factor according to the difference between the data collection timestamp and the current time; Take the historical accuracy rate, supervision intensity score, timeliness decay factor and consistency status of the IP geographic location and platform registration location as input variables; Calculate the comprehensive trust weight value by weighted summation, and the expression is: ； wherein, is a comprehensive trust weight value, is a trustworthiness component, is a timeliness decay factor, is a regulatory strength score, is a geographic trust status value, , , , is a preset weight coefficient; For all data sources Sort by value, select Data streams from data sources exceeding the first threshold are used as the main data channel, while the remaining data enters the verification queue. 4.The method of claim 3, wherein: The commodity service data in the main data channel is input into a multi-language semantic encoder to generate a semantic vector and calculate a cross-language semantic alignment degree, while comparing the logical consistency of key structured fields, specifically including: Extracting the description texts of the same commodity in two different language versions in the main data channel, and inputting them into a pre-trained multi-language semantic encoding model respectively to obtain corresponding high-dimensional semantic vectors; Calculating the cosine similarity between the two semantic vectors as an index of cross-language semantic alignment degree; Extracting the key structured fields corresponding to the texts, including price value, delivery range, and return and exchange period; Judging whether the relative difference of the price value is less than a preset tolerance coefficient, whether the delivery range covers the same area, and whether the return and exchange period is within the allowed deviation range; If the fields all meet the consistency conditions, it is determined that the structured information is consistent; otherwise, it is determined to be inconsistent; The semantic alignment degree and the structured information consistency state are used as the input basis for subsequent consistency score. 5.The method of claim 4, wherein: The consistency score is generated according to the semantic alignment degree and the structured field comparison result, and when the score is lower than a preset threshold, it is marked as abnormal and enters the review process, otherwise it enters the compliance review stage, specifically including: Fusion calculation of the obtained semantic alignment degree and structured information consistency state; A linear weighting method is used to generate a consistency score, the expression is: ; wherein, is a consistency score, is a semantic alignment, is a structured information consistency flag, is a semantic weight coefficient; Will Compared with the second threshold, if If the data is not lower than this threshold, the product / service data is deemed to have cross-language consistency and enters the compliance review stage; If Below the threshold, mark as pending review, generate a diff report containing the diff field and the original text, and push to the human review process. 6.The service data management method of cross-border e-commerce of claim 5, wherein: The regional compliance rule library is called to match the data that passes the consistency check with regulations, and the compliance score is output. Data with a score lower than the threshold is prohibited from synchronization and records risk logs, specifically including: According to the target market country code, determine the jurisdiction where the data is to be published; Retrieve the compliance rule set corresponding to the jurisdiction from the pre-set regional compliance rule library, which includes data privacy, advertising, product access, and consumer protection clauses; Compare item by item whether the commodity service data meets the requirements in each item of the compliance rule set, and count the number of items that meet the requirements; The ratio of the number of items that meet the requirements to the total number of rules is used as the compliance score; The compliance score is compared with the third threshold value. If it is lower than the threshold value, the data is prohibited from participating in cross-border synchronization operation, and the violation rule number and field are recorded in the risk traceability log; If it is not lower than the threshold value, the data is allowed to enter the dispatch queue, and its compliance score is used as one of the input parameters for dispatch priority calculation. 7.The service data management method of cross-border e-commerce of claim 6, wherein: The dispatch priority is generated by combining the data that passes the compliance review with its credibility weight, consistency score, transmission delay, and target market active period, specifically including: Obtain the predicted transmission delay of the data from the edge node to the target market front-end system; Obtain the state identifier of whether the target market is currently in the user access peak period; Normalize the comprehensive credibility weight value, consistency score, and compliance score to obtain the standardized value; Introduce the business conversion potential coefficient, which is determined according to the commodity category and operation state; Calculate the preliminary data value index using a weighted summation method, the expression is: ； wherein is a preliminary data value index, , , are normalized trust weight, consistency score and compliance score, respectively, is a business conversion potential coefficient, , , , is a weight coefficient; Divide the preliminary data value index by the transmission delay, and combine it with the user active period state to adjust the gain, to obtain the dispatch priority; The higher the dispatch priority, the more preferentially the data is executed for cross-border synchronization. 8.The service data management method of cross-border e-commerce of claim 7, wherein: The cross-border data synchronization is performed according to the scheduling priority, and the trusted weight and the consistency evaluation model parameter are dynamically optimized by using user behavior feedback data to form a closed-loop updating mechanism, and specifically comprises: All data passing through the compliance review are sorted according to the final scheduling priority from high to low to form a scheduling queue; Data are taken from the scheduling queue in turn and cross-border transmission is performed to push to the front-end display system or the operation database of the corresponding country region; After the data are put online, the interactive behavior data of the target market users are collected, including the number of clicks, the page stay time, the number of order conversion and the number of complaints; The user feedback score is calculated according to the interactive behavior data to reflect the actual performance of the data quality; The user feedback score is taken as a real label, and regression analysis is performed on the comprehensive trusted weight value and the consistency score to adjust the weight parameters in the trustworthiness model and the consistency evaluation model; The updated model parameters are written into the system configuration for the next round of data processing cycle to realize dynamic iterative optimization of the model parameters. 9.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is characterized in that: The processor executes the computer program to realize the steps of the service data management method of the cross-border e-commerce according to any one of claims 1-8.

10. A computer readable storage medium having stored thereon a computer program, characterized in that: The computer program is executed by the processor to realize the steps of the service data management method of the cross-border e-commerce according to any one of claims 1-8.

Images