Online commodity bogus transaction risk early warning method and system

Abstract

The invention particularly relates to an online commodity bogus transaction risk early warning method and system. The online commodity bogus transaction risk early warning method comprises the following steps: acquiring public commodity information of a network retail platform and e-commerce platform list data in the same time period, and storing the public commodity information and the e-commerce platform list data in a database; standardization processing is carried out, a commodity table and a list commodity table are established, and the monthly sales volume of a commodity list, the lowest monthly sales volume of the list under the same platform category and the highest monthly sales volume of the list under the same-time-period standard category are obtained; judging whether the commodity has a false sales volume risk or not, and performing risk early warning according to a pre-established judgment standard; according to the online commodity bogus transaction risk early-warning method and system, illegal behaviors such as bogus transactions and virtual bid sales volume can be identified in batches, the supervision time cost is reduced, the supervision efficiency is improved, correct reference can be provided for purchase of consumers, compliance guidance can be provided for operators, and data support and factual basis can be provided for market supervision and data analysis.

Description

Technical Field

[0001] The present invention relates to the technical field of e-commerce, and in particular to a method and system for early warning of risks of false online commodity transactions. Background Art

[0002] In recent years, the e-commerce business environment has been continuously optimized, the scale of e-commerce has expanded rapidly, the quality has continued to improve, and the trend of integration and innovation has continued to deepen, which has played an important role in strengthening, optimizing and expanding the digital economy.

[0003] As the times change, more and more entrepreneurs are joining the e-commerce industry. The online store operation model is centered on traffic, and as industry competition gradually intensifies. In order to create hot products, obtain traffic, and attract customers, some merchants make false transactions and use a series of means to make the data look as real as possible to evade platform investigations; some e-commerce platforms condone false transactions for indicators such as user activity, number of merchants settled in, and order growth rate; some e-commerce platforms actively fabricate sales for products in order to promote transactions and obtain service income, resulting in the displayed sales of products being greater than the actual sales. These have affected the normal e-commerce ecology, damaged the legitimate rights and interests of consumers, disrupted the market order of fair transactions, and are not conducive to the management department's understanding of the real market dynamics and the healthy development of the industry.

[0004] Each e-commerce platform has its own rules for dealing with non-compliant behaviors such as fake transactions and fictitious sales, but fake transactions continue despite repeated bans. On the one hand, organized fake transactions evade various audit dimensions of e-commerce platforms. On the other hand, e-commerce platforms cannot effectively identify or restrict such behaviors, and are even motivated to cooperate in displaying false sales. In addition, there are also certain studies on third-party institutions that fake orders, but the research is limited to merchant behavior and ignores platform behavior, and is limited to a single sales indicator and accurate value. Due to the fact that e-commerce platforms only publicly disclose vague sales and comments, the accuracy of the fake order identification model is greatly limited.

[0005] Compliance management is a necessary condition for the sustainable and healthy development of the industry, and is also an important means to safeguard consumer rights and enhance corporate competitiveness. With the increasing importance of e-commerce, establishing a wide-ranging, comprehensive and effective false sales identification and processing system to identify the false transaction risks of merchants and platforms is an urgent problem that must be solved in online retail monitoring, statistics and analysis, which will help regulators to supervise the industry.

[0006] Based on this, the present invention proposes a method and system for early warning of the risk of false online commodity transactions. Summary of the invention

[0007] In order to remedy the defects of the prior art, the present invention provides a simple and efficient online commodity false transaction risk warning method and system.

[0008] The present invention is achieved through the following technical solutions:

[0009] An online commodity false transaction risk early warning method, characterized in that it includes the following steps:

[0010] Step S1: Online product information collection

[0011] Obtain public product information from online retail platforms through big data collection technology, including product ID, product name, product price, product category information, product sales, number of product reviews, number of reviews with pictures, number of payees, product parameters, product brand and product main picture, as well as the store ID, store name, store affiliated company name and store address;

[0012] In addition, obtain the e-commerce platform ranking data of the same period, including ranking categories, product IDs, product names and sales information, and store the collected data in the database;

[0013] Step S2: Standardization

[0014] Standardize the collected product information and create a product table containing the platform, product ID, product name, product brand, product price, product monthly sales volume, product monthly sales, number of payers, total number of product reviews, number of product reviews with pictures, product platform category, product judgment standard category, store ID, store name, company name, company registration province, city, district and county, store location province, city and district and county information, with the primary key being product ID;

[0015] Standardize the list data and create a list product table that includes list categories, product IDs, product names, and monthly sales of products;

[0016] Based on the product ID, the product table is linked to the list product table of the same period to obtain the monthly sales volume of the product list;

[0017] Use platform category association to obtain the lowest monthly sales of the list under the same platform category;

[0018] Use the judgment standard category association to obtain the highest monthly sales of the list under the standard category in the same period;

[0019] In step S2, when the collected commodity information is standardized, the key field processing rules are as follows:

[0020] (1) If the product brand can be directly obtained through collection, the collected product brand is used; if it cannot be directly obtained, the product brand is intercepted from the product parameters using a regular expression; the brand is standardized and processed into the form of "brand English name / brand Chinese name";

[0021] (2) Product price refers to the price of the product, which is the original price of the product minus the price after discount;

[0022] If the platform discloses the price, it will be adopted directly;

[0023] If the product price is a discounted price after participating in a platform activity, the product price will be calculated based on the platform activity algorithm;

[0024] If the price of a product is Pi, and it participates in the platform activity of "40 off for purchases over 300", the real-time transaction price of the product will be Pi*(1-40 / 300).

[0025] (3) Monthly sales volume of goods: Since fraudulent transactions are continuous behaviors over a certain period of time, a sufficient period of time is required for determination. Generally, monthly sales volume is used as a reference, i.e., sales volume over 30 days.

[0026] If the website discloses the cumulative sales volume, the interval between the beginning and the end of the period is t days, and the sales volume is recorded as V0 and Vt respectively, then the corresponding monthly sales volume Q of the product is processed as follows:

[0027]

[0028] If the website discloses monthly sales, it will be used directly;

[0029] If the website discloses annual sales, half-year sales or 7-day sales, the time interval is T days, and the sales at the beginning and end of the period are recorded as M0 and MT respectively;

[0030] When T<30, the monthly sales volume Q of the product adopts the average sales volume, and the processing rules are as follows:

[0031]

[0032] When T>=30, the monthly sales volume Q of the product adopts the difference sales volume, and the processing rules are as follows:

[0033]

[0034] (4) The monthly sales volume of the products on the list is calculated in the same way as the monthly sales volume of the products on the list; the sales volume of the products on the list is calculated in the same way as the sales volume of the products on the list;

[0035] Since the continuity of the products on the list is poor and the time interval is less than 30 days, the average sales volume is used for processing. The processing rules are as follows:

[0036]

[0037] (5) The number of product reviews includes the total number of product reviews and the number of product reviews with pictures;

[0038] The total number of product reviews is the total number of product reviews, and the number of product reviews with pictures is the total number of product reviews with pictures.

[0039] (6) Establish a mapping table between platform categories and judgment standard categories, and treat the platform categories as judgment standard categories. If there is no platform category, treat it as the judgment standard category based on the product name;

[0040] (7) The company's registered place and store location are standardized according to administrative divisions.

[0041] Step S3: Risk warning for fake sales products

[0042] According to the characteristics of false sales data, a model is constructed for analysis of the processed sales data to determine whether there is a risk of false sales of the product, and risk warnings are issued based on the judgment criteria established in advance;

[0043] The judging criteria are as follows:

[0044] If the sales distribution of goods in a store is abnormal, the numerical characteristics of the sales volume of goods in a store are abnormal, the sales fluctuations of the store are abnormal, the gap between the sales volume and the reviews is too large, the gap between the sales volume and the number of payers is too large, the product reviews are abnormal, or the sales volume of goods is significantly different from the sales volume on the same period list, then the sales volume of goods is considered to be inflated and there is a risk of false transactions.

[0045] In step S3, if one of the following conditions is met, it is considered that the sales distribution of goods in the store is abnormal:

[0046] a. The product concentration is too high, that is, the sales of the top 2 products in a store account for a higher proportion of the store's sales than the custom threshold;

[0047] b. The sales volume of products is abnormal. The products in the store within the custom time are sorted from high to low according to sales volume. The sales volume of the product ranked i is V. i , the number of digits of product sales is W i , W i -W i+1 ≥2 and V i / V i+1 Above a custom threshold.

[0048] In step S3, if the sales volume is accurate, the monthly sales volume of the top N (N≥3) products in the same store are the same or similar, in the form of n1*POWER(10,n)+n2, where n1 is an integer from 1 to 9, n≥3, and n2 is an integer with a digit less than n-2. If n1 of the top N products is the same, then the numerical characteristics of the sales volume of the products in the store are considered abnormal.

[0049] In step S3, if one of the following conditions is met, it is considered that the store sales fluctuation is abnormal:

[0050] a. The store's year-on-year and month-on-month sales are both higher than the custom threshold;

[0051] b. Calculate the contribution rate of stores in the region to regional fluctuations based on the month-on-month / year-on-year fluctuations in the proportion of administrative divisions; if the fluctuation in the proportion of the region exceeds the custom threshold, and the contribution rate exceeds the custom threshold;

[0052] c. When the sales ratio of the same brand on competing platforms changes in magnitude (the sales ratio of brand P on platform A and platform B in the base period is PA / PB, and the ratio in the current period is ≥10*PA / PB), the sales ratio on different platforms exceeds the custom threshold.

[0053] In step S3, if the ratio of the monthly sales volume of the product to the cumulative number of product reviews exceeds the custom threshold, it is considered that the gap between the sales volume and the reviews of the product is too large;

[0054] If the ratio of monthly sales volume of a product to the number of cumulative payees exceeds a custom threshold, it is considered that the gap between sales volume and the number of payees is too large;

[0055] If the proportion of comments with images to the total number of comments exceeds a custom threshold, the product reviews are considered abnormal.

[0056] In step S3, if one of the following conditions is met, it is considered that the sales volume of the product is significantly different from the sales volume of the list during the same period:

[0057] a. The ratio of sales volume calculated in the product table to sales volume in the list exceeds the custom threshold;

[0058] b. In the same platform category, the product is not included in the list, but the ratio of the product sales to the lowest monthly sales of the list in the corresponding category exceeds the custom threshold;

[0059] c. Under the same judgment criteria category, the product is not included in the list, but its sales volume is higher than the highest monthly sales volume of the list under this category.

[0060] An online commodity false transaction risk early warning system, comprising:

[0061] The product information collection module is responsible for obtaining public product information on online retail platforms through big data collection technology, including product ID, product name, product price, product category information, product sales, number of product reviews, number of reviews with pictures, number of payees, product parameters, product brand and product main picture, as well as the store ID, store name, store affiliated company name and store address;

[0062] In addition, obtain the e-commerce platform ranking data of the same period, including ranking categories, product IDs, product names and sales information, and store the collected data in the database;

[0063] The product information processing module is responsible for standardizing the collected product information and establishing a product table containing the platform, product ID, product name, product brand, product price, product monthly sales volume, product monthly sales, number of payers, total number of product reviews, number of product reviews with pictures, product platform category, product judgment standard category, store ID, store name, company name, company registration province, city, district and county, store location province, city and district and county information, with the primary key being the product ID;

[0064] Standardize the list data and create a list product table that includes list categories, product IDs, product names, and monthly sales of products;

[0065] Based on the product ID, the product table is linked to the list product table of the same period to obtain the monthly sales volume of the product list;

[0066] Use platform category association to obtain the lowest monthly sales of the list under the same platform category;

[0067] Use the judgment standard category association to obtain the highest monthly sales of the list under the standard category in the same period;

[0068] The false sales risk identification and early warning module is responsible for building a model for analyzing the processed sales data to determine whether the product has a false sales risk and to issue a risk warning based on pre-established judgment criteria.

[0069] A computing device comprising:

[0070] One or more processors, one or more memories, and one or more programs, wherein the one or more programs are stored in the one or more memories and are configured to be executed by the one or more processors, and the one or more programs include instructions for executing any of the above methods.

[0071] A readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described above is implemented.

[0072] The beneficial effects of the present invention are as follows: the online commodity false transaction risk warning method and system can identify false transactions, false sales volume and other violations in batches, reduce the time and economic costs of supervision, improve supervision efficiency, provide correct reference for consumer purchases, provide compliance guidance for operators, and provide data support and factual basis for market supervision and data analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0074] Attached Figure 1 It is a schematic diagram of the risk early warning method for false online commodity transactions of the present invention. DETAILED DESCRIPTION

[0075] In order to enable those skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0076] The online commodity false transaction risk warning method comprises the following steps:

[0077] Step S1: Online product information collection

[0078] Obtain public product information from online retail platforms through big data collection technology, including product ID, product name, product price, product category information, product sales, number of product reviews, number of reviews with pictures, number of payees, product parameters, product brand and product main picture, as well as the store ID, store name, store affiliated company name and store address;

[0079] In addition, obtain the e-commerce platform ranking data of the same period, including ranking categories, product IDs, product names and sales information, and store the collected data in the database;

[0080] Step S2: Standardization

[0081] Standardize the collected product information and create a product table containing the platform, product ID, product name, product brand, product price, product monthly sales volume, product monthly sales, number of payers, total number of product reviews, number of product reviews with pictures, product platform category, product judgment standard category, store ID, store name, company name, company registration province, city, district and county, store location province, city and district and county information, with the primary key being product ID;

[0082] Standardize the list data and create a list product table that includes list categories, product IDs, product names, and monthly sales of products;

[0083] Based on the product ID, the product table is linked to the list product table of the same period to obtain the monthly sales volume of the product list;

[0084] Use platform category association to obtain the lowest monthly sales of the list under the same platform category;

[0085] Use the judgment standard category association to obtain the highest monthly sales of the list under the standard category in the same period;

[0086] In step S2, when the collected commodity information is standardized, the key field processing rules are as follows:

[0087] (1) If the product brand can be directly obtained through collection, the collected product brand is used; if it cannot be directly obtained, the product brand is intercepted from the product parameters using a regular expression; the brand is standardized and processed into the form of "brand English name / brand Chinese name";

[0088] (2) Product price refers to the price of the product, which is the original price of the product minus the price after discount;

[0089] If the platform discloses the price, it will be adopted directly;

[0090] If the product price is a discounted price after participating in a platform activity, the product price will be calculated based on the platform activity algorithm;

[0091] If the price of a product is Pi, and it participates in the platform activity of "40 off for purchases over 300", the real-time transaction price of the product will be Pi*(1-40 / 300).

[0092] (3) Monthly sales volume of goods: Since fraudulent transactions are continuous behaviors over a certain period of time, a sufficient period of time is required for determination. Generally, monthly sales volume is used as a reference, i.e., sales volume over 30 days.

[0093] If the website discloses the cumulative sales volume, the interval between the beginning and the end of the period is t days, and the sales volume is recorded as V0 and Vt respectively, then the corresponding monthly sales volume Q of the product is processed as follows:

[0094]

[0095] If the website discloses monthly sales, it will be used directly;

[0096] If the website discloses annual sales, half-year sales or 7-day sales, the time interval is T days, and the sales at the beginning and end of the period are recorded as M0 and MT respectively;

[0097] When T<30, the monthly sales volume Q of the product adopts the average sales volume, and the processing rules are as follows:

[0098]

[0099] When T>=30, the monthly sales volume Q of the product adopts the difference sales volume, and the processing rules are as follows:

[0100]

[0101] (4) The monthly sales volume of the products on the list is calculated in the same way as the monthly sales volume of the products on the list; the sales volume of the products on the list is calculated in the same way as the sales volume of the products on the list;

[0102] Since the continuity of the products on the list is poor and the time interval is less than 30 days, the average sales volume is used for processing. The processing rules are as follows:

[0103]

[0104] (5) The number of product reviews includes the total number of product reviews and the number of product reviews with pictures;

[0105] The total number of product reviews is the total number of product reviews, and the number of product reviews with pictures is the total number of product reviews with pictures.

[0106] (6) Establish a mapping table between platform categories and judgment standard categories, and treat the platform categories as judgment standard categories. If there is no platform category, treat it as the judgment standard category based on the product name;

[0107] (7) The company's registered place and store location are standardized according to administrative divisions.

[0108] Step S3: Risk warning for fake sales products

[0109] According to the characteristics of false sales data, a model is constructed for analysis of the processed sales data to determine whether there is a risk of false sales of the product, and risk warnings are issued based on the judgment criteria established in advance;

[0110] The judging criteria are as follows:

[0111] If the sales distribution of goods in a store is abnormal, the numerical characteristics of the sales volume of goods in a store are abnormal, the sales fluctuations of the store are abnormal, the gap between the sales volume and the reviews is too large, the gap between the sales volume and the number of payers is too large, the product reviews are abnormal, or the sales volume of goods is significantly different from the sales volume on the same period list, then the sales volume of goods is considered to be inflated and there is a risk of false transactions.

[0112] In step S3, if one of the following conditions is met, it is considered that the sales distribution of goods in the store is abnormal:

[0113] a. The product concentration is too high, that is, the sales of the top 2 products in a store account for a higher proportion of the store's sales than the custom threshold;

[0114] b. The sales volume of products is abnormal. The products in the store within the custom time are sorted from high to low according to sales volume. The sales volume of the product ranked i is V. i , the number of digits of product sales is W i , W i -W i+1 ≥2 and V i / V i+1 Above a custom threshold.

[0115] For example, the sales volume of the product in the store ranked No. 1 is 35654 (5-digit number), so the number of digits of the product sales volume is 5; the sales volume of the product in the store ranked No. 2 is 474 (3-digit number), so the number of digits of the product sales volume is 4, and the difference between the sales volume digits of the product ranked No. 1 is 2, and the product sales ratio in the store is 35654 / 474=75.22. Assuming that the custom threshold is set to 50, 75.22 is greater than 50, then it is determined that the sales volume gap of the product in the store ranked No. 1 is abnormal.

[0116] In step S3, if the sales volume is accurate, the monthly sales volume of the top N (N≥3) products in the same store are the same or similar, in the form of n1*POWER(10,n)+n2, where n1 is an integer from 1 to 9, n≥3, and n2 is an integer with a digit less than n-2. If n1 of the top N products is the same, then the numerical characteristics of the sales volume of the products in the store are considered abnormal.

[0117] In step S3, if one of the following conditions is met, it is considered that the store sales fluctuation is abnormal:

[0118] a. The store's year-on-year and month-on-month sales are both higher than the custom threshold;

[0119] b. Calculate the contribution rate of stores in the region to regional fluctuations based on the month-on-month / year-on-year fluctuations in the proportion of administrative divisions; if the fluctuation in the proportion of the region exceeds the custom threshold, and the contribution rate exceeds the custom threshold;

[0120] c. When the sales ratio of the same brand on competing platforms changes in magnitude (the sales ratio of brand P on platform A and platform B in the base period is PA / PB, and the ratio in the current period is ≥10*PA / PB), the sales ratio on different platforms exceeds the custom threshold.

[0121] In step S3, if the ratio of the monthly sales volume of the product to the cumulative number of product reviews exceeds the custom threshold, it is considered that the gap between the sales volume and the reviews of the product is too large;

[0122] If the ratio of monthly sales volume of a product to the number of cumulative payees exceeds a custom threshold, it is considered that the gap between sales volume and the number of payees is too large;

[0123] If the proportion of comments with images to the total number of comments exceeds a custom threshold, the product reviews are considered abnormal.

[0124] In step S3, if one of the following conditions is met, it is considered that the sales volume of the product is significantly different from the sales volume of the list during the same period:

[0125] a. The ratio of sales volume calculated in the product table to sales volume in the list exceeds the custom threshold;

[0126] b. In the same platform category, the product is not included in the list, but the ratio of the product sales to the lowest monthly sales of the list in the corresponding category exceeds the custom threshold;

[0127] c. Under the same judgment criteria category, the product is not included in the list, but its sales volume is higher than the highest monthly sales volume of the list under this category.

[0128] The online commodity false transaction risk early warning system includes:

[0129] The product information collection module is responsible for obtaining public product information on online retail platforms through big data collection technology, including product ID, product name, product price, product category information, product sales, number of product reviews, number of reviews with pictures, number of payees, product parameters, product brand and product main picture, as well as the store ID, store name, store affiliated company name and store address;

[0130] In addition, obtain the e-commerce platform ranking data of the same period, including ranking categories, product IDs, product names and sales information, and store the collected data in the database;

[0131] The product information processing module is responsible for standardizing the collected product information and establishing a product table containing the platform, product ID, product name, product brand, product price, product monthly sales volume, product monthly sales, number of payers, total number of product reviews, number of product reviews with pictures, product platform category, product judgment standard category, store ID, store name, company name, company registration province, city, district and county, store location province, city and district and county information, with the primary key being the product ID;

[0132] Standardize the list data and create a list product table that includes list categories, product IDs, product names, and monthly sales of products;

[0133] Based on the product ID, the product table is linked to the list product table of the same period to obtain the monthly sales volume of the product list;

[0134] Use platform category association to obtain the lowest monthly sales of the list under the same platform category;

[0135] Use the judgment standard category association to obtain the highest monthly sales of the list under the standard category in the same period;

[0136] The false sales risk identification and early warning module is responsible for building a model for analyzing the processed sales data to determine whether the product has a false sales risk and to issue a risk warning based on pre-established judgment criteria.

[0137] The computer-readable storage medium stores one or more programs, wherein the one or more programs include instructions, and when the instructions are executed by a computing device, the computing device performs any one of the above methods.

[0138] The computing device comprises:

[0139] One or more processors, one or more memories, and one or more programs, wherein the one or more programs are stored in the one or more memories and are configured to be executed by the one or more processors, and the one or more programs include instructions for executing any of the above methods.

[0140] This online commodity false transaction risk warning method and system uses big data collection technology and big data analysis technology to obtain basic information, sales information, comment information, etc. of commodities on sale on the e-commerce platform, build an automated data processing flow, generate standardized processing fields, and construct a model for analysis of the processed sales data, so as to determine whether the commodity has the risk of false transactions or false sales, and issue risk warnings based on pre-established standards.

[0141] The use of this system can help understand the real transaction status of goods, identify false transactions, false sales volume violations, and issue risk warnings. It can not only provide consumers with real references for purchasing goods and guide consumers to consume rationally; it is also conducive to regulating merchants' trading behaviors and maintaining a fair trading market order; it can also provide an objective basis for regulatory authorities to strengthen e-commerce supervision, maintain a fair trading market order, and ensure the steady growth of online transactions.

[0142] The embodiment described above is only one specific implementation of the present invention. Common changes and substitutions made by those skilled in the art within the scope of the technical solution of the present invention should be included in the protection scope of the present invention.

Claims

1. A risk warning method for online commodity fraudulent transactions, characterized by: The following steps are involved: Step S1: Online product information collection Obtain public product information from online retail platforms through big data collection technology, including product ID, product name, product price, product category information, product sales, number of product reviews, number of reviews with pictures, number of payees, product parameters, product brand and product main picture, as well as the store ID, store name, store affiliated company name and store address; In addition, obtain the e-commerce platform ranking data of the same period, including ranking categories, product IDs, product names and sales information, and store the collected data in the database; Step S2: Standardization Standardize the collected product information and create a product table containing the platform, product ID, product name, product brand, product price, product monthly sales volume, product monthly sales, number of payers, total number of product reviews, number of product reviews with pictures, product platform category, product judgment standard category, store ID, store name, company name, company registration province, city, district and county, store location province, city and district and county information, with the primary key being product ID; Standardize the list data and create a list product table that includes list categories, product IDs, product names, and monthly sales of products; Based on the product ID, the product table is linked to the list product table of the same period to obtain the monthly sales volume of the product list; Use platform category association to obtain the lowest monthly sales of the list under the same platform category; Use the judgment standard category association to obtain the highest monthly sales of the list under the standard category in the same period; Step S3: Risk warning for fake sales products According to the characteristics of false sales data, a model is constructed for analysis of the processed sales data to determine whether there is a risk of false sales of the product, and risk warnings are issued based on the judgment criteria established in advance; The judging criteria are as follows: If the sales distribution of goods in a store is abnormal, the numerical characteristics of the sales volume of goods in a store are abnormal, the sales fluctuations of the store are abnormal, the gap between the sales volume and the reviews is too large, the gap between the sales volume and the number of payers is too large, the product reviews are abnormal, or the sales volume of goods is significantly different from the sales volume on the same period list, then the sales volume of goods is considered to be inflated and there is a risk of false transactions.

2. The online commodity fraud transaction risk early warning method according to claim 1 is characterized by: In step S2, when the collected commodity information is standardized, the key field processing rules are as follows: (1) If the product brand can be directly obtained through collection, the collected product brand is used; if it cannot be directly obtained, the product brand is intercepted from the product parameters using a regular expression; the brand is standardized and processed into the form of "brand English name / brand Chinese name"; (2) Product price refers to the price of the product, which is the original price of the product minus the price after discount; If the platform discloses the price, it will be adopted directly; If the product price is a discounted price after participating in a platform activity, the product price will be calculated based on the platform activity algorithm; (3) Monthly sales volume of goods: Since fraudulent transactions are continuous behaviors over a certain period of time, a sufficient period of time is required for determination. Generally, monthly sales volume is used as a reference, i.e., sales volume over 30 days. If the website discloses the cumulative sales volume, there is a time interval of t days between the beginning and the end of the period, and the sales volume is recorded as V0 and V t , then the corresponding product monthly sales volume Q processing rules are as follows: If the website discloses monthly sales, it will be used directly; If the website discloses annual sales, half-year sales or 7-day sales, the time interval is T days, and the sales at the beginning and end of the period are recorded as M0 and M T ; When T<30, the monthly sales volume Q of the product adopts the average sales volume, and the processing rules are as follows: When T>=30, the monthly sales volume Q of the product adopts the difference sales volume, and the processing rules are as follows: (4) The monthly sales volume of products on the list is calculated in the same way as the monthly sales volume of products on the list; the sales volume of products on the list is calculated in the same way as the sales volume of products on the list; Since the continuity of the products on the list is poor and the time interval is less than 30 days, the average sales volume is used for processing. The processing rules are as follows: (5) The number of product reviews includes the total number of product reviews and the number of product reviews with pictures; The total number of product reviews is the total number of product reviews, and the number of product reviews with pictures is the total number of product reviews with pictures. (6) Establish a mapping table between platform categories and judgment standard categories, and treat the platform categories as judgment standard categories. If there is no platform category, treat it as the judgment standard category based on the product name; (7) The company's registered place and store location are standardized according to administrative divisions.

3. The online commodity fraud transaction risk early warning method according to claim 1 is characterized by: In step S3, if one of the following conditions is met, it is considered that the sales distribution of goods in the store is abnormal: a. The product concentration is too high, that is, the sales of the top 2 products in a store account for a higher proportion of the store's sales than the custom threshold; b. The sales volume of products is abnormal. The products in the store within the custom time are sorted from high to low according to sales volume. The sales volume of the product ranked i is V. i , the number of digits of product sales is W i , W i -W i+1 ≥2 and V i / V i+1 Above a custom threshold.

4. The online commodity fraud transaction risk early warning method according to claim 1 is characterized by: In step S3, if the sales volume is accurate, the monthly sales volume of the top N (N≥3) products in the same store are the same or similar, in the form of n1*POWER(10,n)+n2, where n1 is an integer from 1 to 9, n≥3, and n2 is an integer with a digit less than n-2. If n1 of the top N products is the same, then the numerical characteristics of the sales volume of the products in the store are considered abnormal.

5. The online commodity fraud transaction risk early warning method according to claim 1 is characterized by: In step S3, if one of the following conditions is met, it is considered that the store sales fluctuation is abnormal: a. The store's year-on-year and month-on-month sales are both higher than the custom threshold; b. Calculate the contribution rate of stores in the region to regional fluctuations based on the month-on-month / year-on-year fluctuations in the proportion of administrative divisions; if the fluctuation in the proportion of the region exceeds the custom threshold, and the contribution rate exceeds the custom threshold; c. When the sales ratio of the same brand on competing platforms changes in magnitude (the sales ratio of brand P on platform A and platform B in the base period is PA / PB, and the ratio in the current period is ≥10*PA / PB), the sales ratio on different platforms exceeds the custom threshold.

6. The online commodity fraud transaction risk early warning method according to claim 1, characterized in that: In step S3, if the ratio of the monthly sales volume of the product to the cumulative number of product reviews exceeds the custom threshold, it is considered that the gap between the sales volume and the reviews of the product is too large; If the ratio of monthly sales volume of a product to the number of cumulative payees exceeds a custom threshold, it is considered that the gap between sales volume and the number of payees is too large; If the proportion of comments with images to the total number of comments exceeds a custom threshold, the product reviews are considered abnormal.

7. The online commodity fraud transaction risk early warning method according to claim 1 is characterized by: In step S3, if one of the following conditions is met, it is considered that the sales volume of the product is significantly different from the sales volume of the list during the same period: a. The ratio of sales volume calculated in the product table to sales volume in the list exceeds the custom threshold; b. In the same platform category, the product is not included in the list, but the ratio of the product sales to the lowest monthly sales of the list in the corresponding category exceeds the custom threshold; c. Under the same judgment criteria category, the product is not included in the list, but its sales volume is higher than the highest monthly sales volume of the list under this category.

8. An online commodity fraud transaction risk early warning system, characterized by: include: The product information collection module is responsible for obtaining public product information on online retail platforms through big data collection technology, including product ID, product name, product price, product category information, product sales, number of product reviews, number of reviews with pictures, number of payees, product parameters, product brand and product main picture, as well as the store ID, store name, store affiliated company name and store address; In addition, obtain the e-commerce platform ranking data of the same period, including ranking categories, product IDs, product names and sales information, and store the collected data in the database; The product information processing module is responsible for standardizing the collected product information and establishing a product table containing the platform, product ID, product name, product brand, product price, product monthly sales volume, product monthly sales, number of payers, total number of product reviews, number of product reviews with pictures, product platform category, product judgment standard category, store ID, store name, company name, company registration province, city, district and county, store location province, city and district and county information, with the primary key being the product ID; Standardize the list data and create a list product table that includes list categories, product IDs, product names, and monthly sales of products; Based on the product ID, the product table is linked to the list product table of the same period to obtain the monthly sales volume of the product list; Use platform category association to obtain the lowest monthly sales of the list under the same platform category; Use the judgment standard category association to obtain the highest monthly sales of the list under the standard category in the same period; The false sales risk identification and early warning module is responsible for building a model for analyzing the processed sales data to determine whether the product has a false sales risk and to issue a risk warning based on pre-established judgment criteria.

9. A computer-readable storage medium storing one or more programs, characterized in that: The one or more programs include instructions which, when executed by a computing device, cause the computing device to perform any one of the methods according to claims 1 to 7.

10. A computing device, characterized in that :include: One or more processors, one or more memories, and one or more programs, wherein the one or more programs are stored in the one or more memories and are configured to be executed by the one or more processors, and the one or more programs include instructions for executing any of the methods according to claims 1 to 7.

Images