Intelligent order management method and system for cross-border e-commerce platform

Abstract

The invention discloses an intelligent order management method for a cross-border e-commerce platform, and the method comprises the steps: obtaining the order data of cross-border e-commerce platform merchants in a set time period, and extracting good comment orders; extracting a target order according to the signing time, the receiving confirmation time and the evaluation time of the well-commented order; and determining whether the target order is abnormal according to the evaluation text content of the target order, the commodity browsing duration and the commodity browsing frequency, and when the ratio of the number of the abnormal orders to the total number of the orders reaches a second threshold value, marking that the cross-border e-commerce platform merchant is abnormal. Whether the order is abnormal or not is comprehensively evaluated from the aspects of the good reputation number, signing time, good reputation time, evaluation text and the like of the order of the e-commerce platform, then whether the merchant to which the order belongs carries out the click farming behavior or not can be evaluated, meanwhile, the buyer can be evaluated according to the buyer data of the abnormal order, the abnormal order is recognized from multiple aspects, and the user experience is improved. The order abnormity identification accuracy of the e-commerce platform is improved, and the order management range is expanded.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis and management, and particularly relates to an intelligent order management method and system for a cross-border e-commerce platform. Background Art

[0002] With the improvement of the network construction coverage rate, the explosion of Internet users has promoted the booming development of the Internet industry. Among them, the e-commerce platform, as a rapidly developing leading Internet industry, has greatly changed people's clothing, food, housing and transportation. Online shopping has become an essential part of people's daily lives. At the same time, with the improvement of people's living standards, the demand for cross-border goods has gradually increased, thus giving rise to cross-border e-commerce platforms. Cross-border e-commerce refers to an e-commerce platform and an online trading platform where trading entities belonging to different customs territories reach transactions, conduct payment settlements through the e-commerce platform, and deliver goods through cross-border logistics to complete the transactions. As a new form of international trade, cross-border e-commerce networkizes and electronizes traditional international trade, mainly using electronic technology and logistics as means, with commerce as the core, moving traditional sales and shopping channels to the Internet, breaking the tangible and intangible barriers between countries and regions, and having advantages such as reducing intermediate links and saving costs.

[0003] However, during the development of e-commerce, some lawbreakers take advantage of the regulatory loopholes and legal gaps in e-commerce platforms and form a well-defined brushstroke industrial chain with the help of Internet technology. Brushstroke refers to the act of merchants obtaining false completed order quantities in improper ways, such as fabricating orders or colluding with users, without real distribution, in order to obtain improper benefits such as increasing the account praise rate, merchant rating, or product sales volume. The brushstroke behavior pollutes the data of e-commerce platforms, seriously disrupts the market competition order, destroys the business integrity system of e-commerce platforms, and damages the rights and interests of users. The increasingly rampant brushstroke behavior has become a cancer in the development of e-commerce platforms.

[0004] With the popularization of e-commerce platforms and the development of the brushstroke industry, the transaction volume of e-commerce platforms is getting larger and larger. Different management of orders on e-commerce platforms can better promote the completion of orders and the management of merchants. This not only includes the statistics and calculation of order data, but also the abnormal identification of orders, that is, it is also extremely important to identify whether an order belongs to a brushstroke behavior. However, as the brushstroke industrial chain grows and the number of brushstroke orders increases, the difficulty of identification also becomes greater. Therefore, the accurate identification of brushstroke transactions and users with brushstroke behavior has become a necessary and arduous task for each e-commerce platform. Summary of the Invention

[0005] The present invention provides an intelligent order management method and system for a cross-border e-commerce platform, which comprehensively evaluates whether an order is abnormal from perspectives such as the number of positive reviews, receipt time, positive review time, and review text of the e-commerce platform order, and thus can evaluate whether the merchant to which the order belongs conducts brush order behavior. At the same time, the buyer can also be evaluated based on the buyer data of the abnormal order, and the abnormal order can be identified from multiple perspectives, improving the accuracy of identifying abnormal orders on the e-commerce platform and expanding the scope of order management.

[0006] The present invention provides an intelligent order management method for a cross-border e-commerce platform, including:

[0007] Obtain the order data of the merchants on the cross-border e-commerce platform within a set time period; wherein, the order data includes the order quantity and order details;

[0008] When the ratio of the number of positive review orders to the total number in the order data reaches a first threshold, extract the positive review orders in the order data;

[0009] Extract the target orders in the positive review orders according to the receipt time, confirmed receipt time, and review time of the positive review orders;

[0010] Determine whether the target order is abnormal according to the review text content, product browsing duration, and product browsing times of the target order. When the target order is abnormal, mark the target order as an abnormal order;

[0011] Count the number of abnormal orders. When the ratio of the number of abnormal orders to the total number of orders reaches a second threshold, mark the cross-border e-commerce platform merchant as abnormal.

[0012] Further, after the step of counting the number of abnormal orders and marking the cross-border e-commerce platform merchant as abnormal when the ratio of the number of abnormal orders to the total number of orders reaches a second threshold, it further includes:

[0013] Obtain one of the abnormal orders and extract the buyer data in the abnormal order; wherein, the buyer data includes the user registration time, names of all purchased products, and customer service communication situation;

[0014] Evaluate and calculate the buyer data value from four dimensions according to the buyer data; the four dimensions are respectively the relevance of the products purchased by the buyer, the buyer's user registration time, the difference between the order price and the actual price, and the number of customer service communications;

[0015] Calculate the buyer data values of all abnormal orders, and sort all buyers from high to low according to all buyer data values.

[0016] Further, the step of extracting the target order from the favorable review orders according to the receipt time, confirmed receipt time, and evaluation time of the favorable review orders includes:

[0017] Determine the time interval between the receipt time and the confirmed receipt time of the favorable review order;

[0018] When the time interval between the receipt time and the confirmed receipt time is less than the first set value, determine the time interval between the confirmed receipt time and the evaluation time;

[0019] When the time interval between the confirmed receipt time and the evaluation time is less than the second set value, extract this favorable review order as the target order;

[0020] Traverse all the favorable review orders to extract multiple target orders from the favorable review orders.

[0021] Further, the step of determining whether the target order is abnormal according to the evaluation text content, product browsing duration, and product browsing times of the target order, and marking the target order as an abnormal order when the target order is abnormal includes:

[0022] Extract the evaluation pictures of the target order, and determine whether the evaluation pictures are exactly the same as other evaluation pictures; wherein, the other evaluation pictures are pictures of other evaluations of the same product;

[0023] If the evaluation pictures are exactly the same as other evaluation pictures, determine that the target order is abnormal;

[0024] If the evaluation pictures are not exactly the same as other evaluation pictures, extract the evaluation text content of the target order, compare the evaluation text content with a preset favorable review keyword library to extract the keywords that appear in the favorable review keyword library in the evaluation text;

[0025] Count the number of characters of the keywords, and determine the ratio of the number of characters of the keywords to the total number of characters of the evaluation text;

[0026] When the ratio of the number of characters of the keywords to the total number of characters of the evaluation text exceeds the first ratio, determine the buyer browsing data before placing the order for the target order; wherein, the buyer browsing data includes browsing duration and browsing times;

[0027] Determine whether the target order is abnormal according to the buyer browsing data;

[0028] When the target order is abnormal, mark the target order as an abnormal order, and traverse all the target orders to extract multiple abnormal orders from the target orders.

[0029] Further, the step of determining whether the target order is abnormal according to the buyer's browsing data includes:

[0030] When the browsing duration is less than the set duration, it is determined that the target order is abnormal;

[0031] When the browsing duration is greater than or equal to the set duration and the number of browsing times is less than the set number of times, it is determined that the target order is abnormal.

[0032] Further, the step of evaluating and calculating the buyer data value from four dimensions according to the buyer data includes:

[0033] Calculating the buyer data value of the first dimension according to the relevance of all the goods purchased by the buyer user;

[0034] Determining the registration time of the buyer user, and obtaining the buyer data value of the second dimension from a preset score query table according to the registration time of the buyer user;

[0035] Calculating the buyer value of the third dimension according to the difference between the prices of all the orders of the buyer user and their actual prices;

[0036] Determining the number of times of customer service communication for all the orders of the buyer user, and obtaining the buyer data value of the fourth dimension from a preset score query table according to the ratio of the number of times of customer service communication to the number of all orders;

[0037] Calculating the buyer data value according to the buyer data value of the first dimension, the buyer data value of the second dimension, the buyer value of the third dimension and the buyer value of the fourth dimension. The calculation formula is: buyer data value = buyer data value of the first dimension × 0.3 + buyer data value of the second dimension × 0.1 + buyer value of the third dimension × 0.4 + buyer value of the fourth dimension × 0.2.

[0038] Further, the step of calculating the buyer data value of the first dimension according to the relevance of all the goods purchased by the buyer user includes:

[0039] Determining the category of each commodity purchased by the buyer user, and counting the number of commodities under each category;

[0040] When the number of commodities under a category is less than the set number, mark the category;

[0041] Counting the ratio of the marked categories to the total categories, and obtaining the buyer data value of the first dimension from a preset score query table according to the ratio.

[0042] Further, the step of calculating the buyer value of the third dimension according to the difference between the prices of all the orders of the buyer user and their actual prices includes:

[0043] Determine the difference between the order price and the actual price in an order; wherein, the actual price is the average of the market prices of the commodity.

[0044] Query a score value of an order in a preset price scoring table according to the ratio of the difference to the order price.

[0045] Take the average of the score values of all orders of the buyer user as the third-dimensional buyer data value.

[0046] The present invention also provides an order intelligent management system for a cross-border e-commerce platform, including:

[0047] An acquisition module, configured to acquire order data of merchants on a cross-border e-commerce platform within a set time period; wherein, the order data includes the order quantity and order details.

[0048] A first extraction module, configured to extract the favorable comment orders in the order data when the ratio of the number of favorable comment orders to the total number in the order data reaches a first threshold.

[0049] A second extraction module, configured to extract target orders in the favorable comment orders according to the receipt time, confirmed receipt time, and evaluation time of the favorable comment orders.

[0050] A determination module, configured to determine whether the target order is abnormal according to the evaluation text content, product browsing duration, and product browsing times of the target order, and mark the target order as an abnormal order when the target order is abnormal.

[0051] A marking module, configured to count the number of abnormal orders, and mark the merchant on the cross-border e-commerce platform as abnormal when the ratio of the number of abnormal orders to the total order quantity reaches a second threshold.

[0052] The present invention also provides a computer device, including a memory and a processor, where the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0053] The present invention also provides a computer-readable storage medium, on which a computer program is stored, and the computer program implements the steps of the above method when executed by a processor.

[0054] The beneficial effects of the present invention are:

[0055] The present invention obtains order data, extracts favorable comment orders according to the ratio of the number of favorable comment orders to the total number, extracts target orders according to the signing time, confirmed receipt time, and evaluation time of the favorable comment orders, determines whether the target orders are abnormal based on the evaluation text content, product browsing duration, and product browsing times of the target orders, and further determines whether the merchants on the e-commerce platform are abnormal, that is, whether they have engaged in order brushing behavior, based on the ratio of the number of abnormal orders to the total number of orders, so as to identify abnormal orders from multiple perspectives, improve the accuracy of order anomaly identification on the e-commerce platform, and expand the scope of order management. In addition, through the buyer data of abnormal orders, it is also possible to identify whether the buyer is an abnormal number, that is, to determine whether the buyer is a brush order number, and to achieve the identification and management of buyers through abnormal orders. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 It is a schematic flowchart of the method according to an embodiment of the present invention.

[0057] Figure 2 It is a schematic structural diagram of the device according to an embodiment of the present invention.

[0058] Figure 3 It is a schematic internal structure diagram of a computer device according to an embodiment of the present invention.

[0059] The implementation, functional features, and advantages of the objectives of the present invention will be further described in conjunction with the embodiments with reference to the accompanying drawings. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0060] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0061] As Figure 1 shown, the present invention provides an order intelligent management method for a cross-border e-commerce platform, including:

[0062] S1. Obtain the order data of merchants on the cross-border e-commerce platform within a set time period; wherein, the order data includes the number of orders and order details;

[0063] S2. When the ratio of the number of favorable comment orders in the order data to the total number reaches a first threshold, extract the favorable comment orders in the order data;

[0064] S3. Extract the target orders in the favorable comment orders according to the signing time, confirmed receipt time, and evaluation time of the favorable comment orders;

[0065] S4. Determine whether the target orders are abnormal according to the evaluation text content, product browsing duration, and product browsing times of the target orders, and when the target orders are abnormal, mark the target orders as abnormal orders;

[0066] S5. Count the number of the abnormal orders. When the ratio of the number of the abnormal orders to the total number of orders reaches a second threshold, mark the merchant of the cross-border e-commerce platform as abnormal.

[0067] As described in the above steps S1 - S5, obtain the order data of the cross-border e-commerce platform within a set time period (such as one month). The order data includes the order quantity and order details. The order details specifically include evaluation content, shipping time, buyer browsing time, logistics data, etc. These data are helpful for identifying whether an order is abnormal. The evaluation of an order includes good reviews, medium reviews, and bad reviews. When the number of orders with good reviews within one month accounts for 4 / 5 of the total number of orders (the first threshold, which can be adjusted according to specific needs and is not limited here, the same below), extract the orders with good reviews. There are multiple such orders. Among the orders with good reviews, target orders can be extracted according to the difference between the confirmed receipt time and the evaluation time based on their signing times. There are multiple target orders. For each target order, it can be determined whether it is an abnormal order according to its evaluation text content, product browsing duration, and product browsing times to complete the identification of all target orders. Finally, when the number of abnormal orders within one month accounts for 80% (the second threshold) of the total number of orders, it can be determined that the merchant is abnormal, that is, there is a behavior of brushing orders.

[0068] In one embodiment, after the step of counting the number of the abnormal orders and marking the merchant of the cross-border e-commerce platform as abnormal when the ratio of the number of the abnormal orders to the total number of orders reaches the second threshold, the following steps are further included:

[0069] S6. Obtain one of the abnormal orders and extract the buyer data in the abnormal order; wherein, the buyer data includes user registration time, names of all purchased products, and customer service communication situation.

[0070] S7. Evaluate and calculate the buyer data value from four dimensions according to the buyer data; wherein, the four dimensions are respectively the relevance of the products purchased by the buyer, the user registration time of the buyer, the difference between the order price and the actual price, and the number of customer service communications.

[0071] S8. Calculate the buyer data values of all abnormal orders and sort all buyers from high to low according to all the buyer data values.

[0072] As described in the above steps S6 - S8, after obtaining the abnormal orders, one of the abnormal orders is acquired, and the buyer data in the abnormal order is extracted. The buyer data includes the user registration time, all the names of the purchased goods, and the customer service communication situation. These data are helpful for identifying whether the buyer user is abnormal. According to the buyer data, the buyer data value can be evaluated and calculated from four dimensions (the relevance of the goods purchased by the buyer, the buyer user registration time, the difference between the order price and the actual price, and the number of customer service communications). The specific calculation method will be described in detail later. After obtaining the buyer data value, the buyers are sorted from high to low according to the buyer data value. Among them, the higher the buyer data value, the higher the probability of the buyer being abnormal. Therefore, the probability of the buyer being abnormal (the abnormal ratio indicates that the buyer user is a brush order number) can be sorted from high to low according to the sorting situation. It is also possible to select the top 20 buyers or those with a set buyer data value or above for specific viewing and warning processing.

[0073] In one embodiment, the step of extracting the target order in the favorable comment order according to the receipt time, confirmed receipt time, and evaluation time of the favorable comment order includes:

[0074] S31. Determine the time interval between the receipt time and the confirmed receipt time of the favorable comment order;

[0075] S32. When the time interval between the receipt time and the confirmed receipt time is less than the first set value, determine the time interval between the confirmed receipt time and the evaluation time;

[0076] S33. When the time interval between the confirmed receipt time and the evaluation time is less than the second set value, extract this favorable comment order as the target order;

[0077] S34. Traverse all the favorable comment orders to extract multiple target orders from the favorable comment orders.

[0078] As described in the above steps S31 - S34, for a favorable comment order, determine the time interval between the receipt time and the confirmed receipt time of this favorable comment order. For example, if the receipt time is 12:00 on November 7, 2024, and the confirmed receipt time is 18:00 on November 7, 2024, then the time interval is 6 hours, which is less than 24 hours (the first set value. If it is greater than the first set value, then evaluate the next order). Further, determine the time interval between the confirmed receipt time and the evaluation time. For example, if the evaluation time is 18:30 on November 7, 2024, then the time interval is 30 minutes, which is less than 4 hours (the second set value. If it is greater than the second set value, then evaluate the next order). Extract this favorable comment order as the target order. After traversing all the favorable comment orders, multiple target orders are extracted from the favorable comment orders.

[0079] In one embodiment, the step of determining whether the target order is abnormal according to the evaluation text content, product browsing duration, and product browsing times of the target order, and marking the target order as an abnormal order when the target order is abnormal includes:

[0080] S41. Extract the evaluation pictures of the target order, and determine whether the evaluation pictures are exactly the same as other evaluation pictures; wherein, the other evaluation pictures are pictures of other evaluations of the same product;

[0081] S42. If the evaluation pictures are exactly the same as other evaluation pictures, determine that the target order is abnormal;

[0082] S43. If the evaluation pictures are not exactly the same as other evaluation pictures, extract the evaluation text content of the target order, and compare the evaluation text content with a preset positive comment keyword library to extract the keywords in the evaluation text that appear in the positive comment keyword library;

[0083] S44. Count the number of characters of the keywords, and determine the ratio of the number of characters of the keywords to the total number of characters of the evaluation text;

[0084] S45. When the ratio of the number of characters of the keywords to the total number of characters of the evaluation text exceeds a first ratio, determine the buyer browsing data before placing the target order; wherein, the buyer browsing data includes browsing duration and browsing times;

[0085] S46. Determine whether the target order is abnormal according to the buyer browsing data;

[0086] S47. When the target order is abnormal, mark the target order as an abnormal order, and traverse all target orders to extract multiple abnormal orders from the target orders.

[0087] As described in the above steps S41 - S47, for a target order, extract the evaluation pictures of the target order. When the evaluation pictures are exactly the same as other evaluation pictures (pictures of other evaluations of the goods in this order) (each picture can be found in other evaluations), it is very likely that the order is not a real order. Therefore, directly determine that the target order is abnormal. When the evaluation pictures are not exactly the same as other evaluation pictures (if there is one picture that cannot be found in other evaluations, it can be considered not exactly the same), extract the evaluation text content of the target order, and compare the evaluation text content with a preset good review keyword library (many keywords used in template evaluations are stored in this database). When the number of keywords compared in the evaluation text content accounts for 2 / 3 (the first ratio) of the total number of words in the entire text, it is considered that the evaluation text is a copied or template - edited text, and this order may be an abnormal order, but further determination is still required, that is, determine the browsing data of the buyer before placing the order, and determine whether the target order is abnormal based on the buyer's browsing data. The specific evaluation steps will be described in detail later. When it is determined that the target order is abnormal, mark this order as an abnormal order. After traversing all target orders, multiple abnormal orders are extracted from the target orders.

[0088] In one embodiment, the step of determining whether the target order is abnormal according to the buyer's browsing data includes:

[0089] S461. When the browsing duration is less than the set duration, determine that the target order is abnormal;

[0090] S462. When the browsing duration is greater than or equal to the set duration and the number of browsing times is less than the set number of times, determine that the target order is abnormal.

[0091] As described in the above steps S461 - S462, when the browsing duration is less than 3 minutes (the set duration, if it is greater than the set duration, the order is normal and the evaluation of the next order is carried out), determine that the target order is abnormal; when the browsing duration is greater than 3 minutes (the set duration) and the number of browsing times is less than 3 times (the set number of times, if it is greater than the set number of times, the order is normal and the evaluation of the next order is carried out), determine that the target order is abnormal.

[0092] In one embodiment, the step of evaluating and calculating the buyer data value from four dimensions according to the buyer data includes:

[0093] S71. Calculate the buyer data value of the first dimension according to the relevance of all the goods purchased by the buyer user;

[0094] S72. Determine the registration time of the buyer user, and obtain the buyer data value of the second dimension according to the registration time of the buyer user in a preset score query table;

[0095] S73. Calculate the buyer value of the third dimension based on the difference between the prices of all orders of the buyer user and their actual prices;

[0096] S74. Determine the number of times of customer service communication for all orders of the buyer user, and obtain the buyer data value of the fourth dimension from a preset score query table according to the ratio of the number of times of customer service communication to the number of all orders;

[0097] S75. Calculate the buyer data value according to the buyer data value of the first dimension, the buyer data value of the second dimension, the buyer value of the third dimension, and the buyer data value of the fourth dimension. The calculation formula is: buyer data value = buyer data value of the first dimension × A + buyer data value of the second dimension × B + buyer value of the third dimension × C + buyer data value of the fourth dimension × D, and A + B + C + D = 1.

[0098] As described in the above steps S71 - S75, a score query table is preset in advance. The score query table stores the registration time interval of the buyer user and its corresponding score (which can be adjusted), the numerical interval of the ratio of the number of times of customer service communication to the number of orders and its corresponding score (which can be adjusted), the numerical interval of the ratio of the marked category to the total category and its corresponding score (which can be adjusted), and the numerical interval of the ratio of the price difference to the order price and its corresponding score (which can be adjusted).

[0099] The buyer data value of the first dimension can be calculated according to the relevance of all the products purchased by the buyer user. The specific calculation process will be described in detail later. Determine the registration time of the buyer user, and obtain the buyer data value of the second dimension from the preset score query table according to the registration time. For example, if the registration time is within one month, the score is 100 points; within 3 months, the score is 80 points; within half a year, the score is 60 points; within one year, the score is 40 points; and more than one year, the score is 0 points. Calculate the buyer value of the third dimension according to the difference between the actual prices of all orders and the order products of the buyer user. The specific calculation process will be described in detail later. Determine the number of times of customer service communication for all orders of the buyer user, and the buyer data value of the fourth dimension can be queried from the score query table according to the ratio of this number to the number of all orders of the buyer. Finally, different weights A, B, C, D are assigned to the buyer data value of the first dimension, the buyer data value of the second dimension, the buyer value of the third dimension, and the buyer data value of the fourth dimension according to their importance, and A + B + C + D = 1. In the present invention, 0.3, 0.1, 0.4, and 0.2 are respectively assigned. Finally, the calculation formula for the buyer data value is: buyer data value = buyer data value of the first dimension × 0.3 + buyer data value of the second dimension × 0.1 + buyer value of the third dimension × 0.4 + buyer data value of the fourth dimension × 0.2.

[0100] In one embodiment, the step of calculating the buyer data value of the first dimension according to the relevance of all the products purchased by the buyer user includes:

[0101] S711. Determine the category of each item purchased by the buyer user, and count the number of items in each category;

[0102] S712. When the number of items in a category is less than the set quantity, mark that category;

[0103] S713. Count the ratio of the marked categories to the total number of categories, and obtain the first - dimension buyer data value from the preset score query table according to the ratio.

[0104] As described in the above steps S711 - S713, determine the category of each item purchased by the buyer user. Eventually, multiple categories and the number of items in each category can be obtained. When the number of items in a category is less than 5 (set quantity), then that category may be abnormal, mark that category, to determine the ratio of the marked categories to the total number of categories, and then query to obtain the first - dimension buyer data value.

[0105] In one embodiment, the step of calculating the third - dimension buyer value according to the difference between the order price and the actual price of all orders of the buyer user includes:

[0106] S731. Determine the difference between the order price and the actual price in an order; where the actual price is the average of the market prices of the item.

[0107] S732. Query and obtain the score value of an order from the preset price score table according to the ratio of the difference to the order price;

[0108] S733. Take the average of the score values of all orders of the buyer user as the third - dimension buyer data value.

[0109] As described in the above steps S731 - S733, determine the difference between the order price and the actual price in an order. The actual price is the average price of the item on the market (calculate by taking the prices of the item from 20 different merchants). According to the difference, the score value of an order can be obtained from the score query table. Traverse all orders, then the score values of all orders can be obtained, sum them up and take the average to obtain the average value as the third - dimension buyer data value.

[0110] As Figure 2 shown, the present invention also provides an order intelligent management system for a cross - border e - commerce platform, including:

[0111] An acquisition module 1, configured to acquire the order data of merchants on the cross - border e - commerce platform within a set time period; where the order data includes the order quantity and order details;

[0112] A first extraction module 2, configured to extract the good - review orders in the order data when the ratio of the number of good - review orders to the total number in the order data reaches a first threshold;

[0113] A second extraction module 3, configured to extract target orders from the favorable comment orders according to the signing time, confirmed receipt time, and evaluation time of the favorable comment orders;

[0114] A determination module 4, configured to determine whether the target order is abnormal according to the evaluation text content, product browsing duration, and product browsing times of the target order, and when the target order is abnormal, mark the target order as an abnormal order;

[0115] A marking module 5, configured to count the number of abnormal orders, and when the ratio of the number of abnormal orders to the total number of orders reaches a second threshold, mark the cross-border e-commerce platform merchant as abnormal.

[0116] In one embodiment, it further includes:

[0117] A third extraction module 6, configured to obtain one of the abnormal orders and extract the buyer data in the abnormal order; wherein, the buyer data includes the user registration time, all product names purchased, and customer service communication status;

[0118] A calculation module 7, configured to evaluate and calculate the buyer data value from four dimensions according to the buyer data; wherein, the four dimensions are respectively the relevance of the products purchased by the buyer, the buyer's user registration time, the difference between the order price and the actual price, and the number of customer service communications;

[0119] A sorting module 8, configured to calculate the buyer data values of all abnormal orders and sort all buyers from high to low according to all the buyer data values.

[0120] In one embodiment, the second extraction module 3 includes:

[0121] A first time interval determination unit, configured to determine the time interval between the signing time and the confirmed receipt time of the favorable comment order;

[0122] A second time interval determination unit, configured to determine the time interval between the confirmed receipt time and the evaluation time when the time interval between the signing time and the confirmed receipt time is less than a first set value;

[0123] A target order extraction unit, configured to extract the favorable comment order as a target order when the time interval between the confirmed receipt time and the evaluation time is less than a second set value;

[0124] A target order traversal unit, configured to traverse all favorable comment orders to extract multiple target orders from the favorable comment orders.

[0125] In one embodiment, the determination module 4 includes:

[0126] A judgment unit, configured to extract the evaluation pictures of the target order and determine whether the evaluation pictures are exactly the same as other evaluation pictures; wherein, the other evaluation pictures are pictures of other evaluations for the same commodity;

[0127] A first anomaly determination unit, configured to determine that the target order is abnormal when the evaluation pictures are exactly the same as other evaluation pictures;

[0128] A comparison unit, configured to, when the evaluation pictures are not exactly the same as other evaluation pictures, extract the evaluation text content of the target order, compare the evaluation text content with a preset good review keyword library, so as to extract the keywords that appear in the good review keyword library in the evaluation text;

[0129] A word count unit, configured to count the number of words of the keywords and determine the ratio of the number of words of the keywords to the total number of words of the evaluation text;

[0130] A browsing data determination unit, configured to, when the ratio of the number of words of the keywords to the total number of words of the evaluation text exceeds a first ratio, determine the buyer browsing data before placing the order for the target order; wherein, the buyer browsing data includes browsing duration and browsing times;

[0131] A second anomaly determination unit, configured to determine whether the target order is abnormal according to the buyer browsing data;

[0132] An abnormal order extraction unit, configured to, when the target order is abnormal, mark the target order as an abnormal order, traverse all target orders, so as to extract multiple abnormal orders from the target orders.

[0133] In one embodiment, the second anomaly determination unit includes:

[0134] A first determination subunit, configured to determine that the target order is abnormal when the browsing duration is less than a set duration;

[0135] A second determination subunit, configured to determine that the target order is abnormal when the browsing duration is greater than or equal to the set duration and the browsing times are less than the set times.

[0136] In one embodiment, the calculation module 7 includes:

[0137] A first calculation unit, configured to calculate a first-dimension buyer data value according to the relevance of all commodities purchased by the buyer user;

[0138] A first query unit, configured to determine the registration time of the buyer user, and obtain a second-dimension buyer data value from a preset score query table according to the registration time of the buyer user;

[0139] A second calculation unit for calculating a third - dimension buyer value according to the difference between the total order price and the actual price of the buyer user;

[0140] A second query unit for determining the number of customer service communications for all orders of the buyer user, and obtaining a fourth - dimension buyer data value from a preset score query table according to the ratio of the number of customer service communications to the number of all orders;

[0141] A score calculation unit for calculating a buyer data value according to the first - dimension buyer data value, the second - dimension buyer data value, the third - dimension buyer value, and the fourth - dimension buyer value. The calculation formula is: buyer data value = first - dimension buyer data value×0.3 + second - dimension buyer data value×0.1 + third - dimension buyer value×0.4 + fourth - dimension buyer value×0.2.

[0142] In one embodiment, the first calculation unit includes:

[0143] A quantity statistics subunit for determining the category of each product purchased by the buyer user and counting the quantity of products in each category;

[0144] A category marking subunit for marking a category when the quantity of products in a category is less than a set quantity;

[0145] A first query subunit for counting the ratio of the marked categories to the total number of categories and obtaining the first - dimension buyer data value from a preset score query table according to the ratio.

[0146] In one embodiment, the second calculation unit includes:

[0147] A difference determination subunit for determining the difference between the order price and the actual price in an order; where the actual price is the average of the market prices of the product;

[0148] A second query subunit for querying and obtaining a score value of an order from a preset price score table according to the ratio of the difference to the order price;

[0149] A result subunit for taking the average of the score values of all orders of the buyer user as the third - dimension buyer data value.

[0150] The above - mentioned modules, units, and subunits are all used to correspondingly execute the respective steps in the order intelligent management method for the cross - border e - commerce platform. Their specific implementation manners refer to those described in the method embodiments above and will not be elaborated here.

[0151] As Figure 3 shown, the present invention also provides a computer device, which may be a server, and its internal structure may be as Figure 3As shown. The computer device includes a processor, a memory, a network interface, and a database connected via a system bus. Among them, the processor of the computer design 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, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store all the data required for the process of the order intelligent management method for the cross-border e-commerce platform. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, it implements the order intelligent management method for the cross-border e-commerce platform.

[0152] Those skilled in the art can understand that Figure 3 the structure shown in is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied.

[0153] An embodiment of this application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements any one of the above-mentioned order intelligent management methods for the cross-border e-commerce platform.

[0154] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to the memory, storage, database, or other media provided in this application and used in the embodiments can include non-volatile and / or volatile memories. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0155] It should be noted that in this text, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent in such a process, device, article or method. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of another identical element in the process, device, article or method including such element.

[0156] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the content of the specification and drawings of the present invention, or directly or indirectly applied in other related technical fields, shall be equally included in the patent protection scope of the present invention.

Claims

1. An intelligent order management method for cross-border e-commerce platforms, characterized in that, Including: Obtain the order data of merchants on the cross-border e-commerce platform within a set time period; wherein, the order data includes the order quantity and order details; When the ratio of the number of good-review orders to the total number in the order data reaches the first threshold, extract the good-review orders in the order data; Extract the target orders in the good-review orders according to the receipt time, confirmed receipt time, and evaluation time of the good-review orders; Determine whether the target order is abnormal according to the evaluation text content, product browsing duration, and product browsing times of the target order. When the target order is abnormal, mark the target order as an abnormal order; Count the number of abnormal orders. When the ratio of the number of abnormal orders to the total number of orders reaches the second threshold, mark the cross-border e-commerce platform merchant as abnormal.

2. The order intelligent management method for cross-border e-commerce platforms according to claim 1, characterized in that After the step of counting the number of abnormal orders and marking the cross-border e-commerce platform merchant as abnormal when the ratio of the number of abnormal orders to the total number of orders reaches the second threshold, it further includes: Obtain one of the abnormal orders and extract the buyer data in the abnormal order; wherein, the buyer data includes the user registration time, names of all purchased products, and customer service communication situation; Evaluate and calculate the buyer data value from four dimensions according to the buyer data; wherein, the four dimensions are the relevance of the products purchased by the buyer, the buyer's user registration time, the difference between the order price and the actual price, and the number of customer service communications; Calculate the buyer data values of all abnormal orders and sort all buyers from high to low according to all buyer data values.

3. The order intelligent management method for cross-border e-commerce platforms according to claim 1, characterized in that, The step of extracting the target orders in the good-review orders according to the receipt time, confirmed receipt time, and evaluation time of the good-review orders includes: Determine the time interval between the receipt time and the confirmed receipt time of the good-review order; When the time interval between the receipt time and the confirmed receipt time is less than the first set value, determine the time interval between the confirmed receipt time and the evaluation time; When the time interval between the confirmed receipt time and the evaluation time is less than the second set value, extract the good-review order as the target order; Traverse all good-review orders to extract multiple target orders from the good-review orders.

4. The order intelligent management method for cross-border e-commerce platforms according to claim 3, characterized in that The step of determining whether the target order is abnormal according to the evaluation text content, product browsing duration, and product browsing times of the target order and marking the target order as an abnormal order when the target order is abnormal includes: Extract the evaluation pictures of the target order and determine whether the evaluation pictures are exactly the same as other evaluation pictures; wherein, the other evaluation pictures are pictures of other evaluations of the same product; If the evaluation pictures are exactly the same as other evaluation pictures, determine that the target order is abnormal; If the evaluation pictures are not exactly the same as other evaluation pictures, extract the evaluation text content of the target order, compare the evaluation text content with a preset good-review keyword library to extract the keywords that appear in the good-review keyword library in the evaluation text; Count the number of words of the keywords and determine the ratio of the number of words of the keywords to the total number of words of the evaluation text; When the ratio of the number of characters of the keyword to the total number of characters of the evaluation text exceeds the first ratio, determine the buyer browsing data before placing the target order; wherein, the buyer browsing data includes browsing duration and browsing times; Determine whether the target order is abnormal according to the buyer browsing data; When the target order is abnormal, mark the target order as an abnormal order, and traverse all target orders to extract multiple abnormal orders from the target orders.

5. The order intelligent management method for cross-border e-commerce platforms according to claim 4, characterized in that, The step of determining whether the target order is abnormal according to the buyer browsing data includes: When the browsing duration is less than the set duration, determine that the target order is abnormal; When the browsing duration is greater than or equal to the set duration and the browsing times is less than the set times, determine that the target order is abnormal.

6. The order intelligent management method for cross-border e-commerce platforms according to claim 2, characterized in that The step of evaluating and calculating the buyer data value from four dimensions according to the buyer data includes: Calculate the buyer data value of the first dimension according to the relevance of all the goods purchased by the buyer user; Determine the registration time of the buyer user, and obtain the buyer data value of the second dimension from a preset score query table according to the registration time of the buyer user; Calculate the buyer value of the third dimension according to the difference between the order price and the actual price of all orders of the buyer user; Determine the number of customer service communications for all orders of the buyer user, and obtain the buyer data value of the fourth dimension from a preset score query table according to the ratio of the number of customer service communications to the number of all orders; Calculate the buyer data value according to the buyer data value of the first dimension, the buyer data value of the second dimension, the buyer value of the third dimension, and the buyer data value of the fourth dimension. The calculation formula is: buyer data value = buyer data value of the first dimension × 0.3 + buyer data value of the second dimension × 0.1 + buyer value of the third dimension × 0.4 + buyer data value of the fourth dimension × 0.

2.

7. The order intelligent management method for cross-border e-commerce platforms according to claim 6, characterized in that, The step of calculating the buyer data value of the first dimension according to the relevance of all the goods purchased by the buyer user includes: Determine the category of each commodity purchased by the buyer user, and count the number of commodities in each category; When the number of commodities in a category is less than the set number, mark the category; Count the ratio of the marked categories to the total categories, and obtain the buyer data value of the first dimension from a preset score query table according to the ratio.

8. The order intelligent management method for cross-border e-commerce platforms according to claim 6, characterized in that The step of calculating the buyer value of the third dimension according to the difference between the order price and the actual price of all orders of the buyer user includes: Determine the difference between the order price and the actual price in an order; wherein, the actual price is the average of the market prices of the commodity; Query and obtain the score value of an order from a preset price scoring table according to the ratio of the difference to the order price; Take the average value of the score values of all orders of the buyer user as the buyer data value of the third dimension.

9. An intelligent order management system for cross-border e-commerce platforms, characterized in that, Include: An acquisition module for acquiring the order data of merchants on the cross-border e-commerce platform within a set time period; wherein, the order data includes the order quantity and order details; A first extraction module for extracting the good review orders in the order data when the ratio of the number of good review orders to the total number in the order data reaches the first threshold; A second extraction module, configured to extract target orders from the favorable comment orders according to the receipt time, confirmed receipt time, and evaluation time of the favorable comment orders; A determination module, configured to determine whether the target orders are abnormal according to the evaluation text content, product browsing duration, and product browsing times of the target orders, and when the target orders are abnormal, mark the target orders as abnormal orders; A marking module, configured to count the number of the abnormal orders, and when the ratio of the number of the abnormal orders to the total number of orders reaches a second threshold, mark the cross-border e-commerce platform merchant as abnormal.

10. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, the steps of the method according to any one of claims 1 to 8 are implemented.

Images