Method and system for detecting brushing behavior of e-commerce platform

Abstract

The invention discloses a method and system for detecting a brushing behavior of an e-commerce platform. The method comprises a step of acquiring original data of a plurality of commodities, and calculating characteristic rate values according to the original data, wherein the plurality of commodities comprises brushing commodities and non-brushing commodities, and the characteristic rate values comprise a silent conversion rate, an order consultation rate, a traffic conversion rate, a transaction conversion, an order payment rate, a collection rate, a time difference of commodity reception, abrusher proportion and a stay time at a shop, a step of normalizing the characteristic rate values and converting the characteristic rate values into values whose format is matched with a support vector machine classification method, and obtaining characteristic values, a step of processing characteristic values of the brushing commodities and the characteristic values of the non-brushing commodities by using the support vector machine classification method and obtaining an optimal training model, a step of calculating a characteristic value of target data and obtaining a target characteristic value, and a step of calculating brushing probability of the target commodity according to the optimal training model and the target characteristic value. The optimal training model is obtained by calculating the characteristic rate values, the brushing probability of the target commodity is calculated, and intuitive reference data is provided to a user.

Description

technical field [0001] The present invention relates to the field of e-commerce, in particular to a method and system for detecting swiping behavior on an e-commerce platform. Background technique [0002] With the rapid development of the e-commerce industry, online shopping has gradually become a new way of life, and the competition in the e-commerce industry has also intensified. Driven by interests, e-commerce behavior has gradually become the "hidden rules" of e-commerce platforms. The e-commerce business is developing, and various constraints and regulations are not perfect, so the behavior of swiping orders on the e-commerce platform is relatively serious. [0003] For order-swiping behavior, JD.com's anti-swipe-swiping system makes statistics from multiple dimensions such as orders, products, users, and logistics, and calculates different characteristic values ​​in each dimension, which can more accurately identify malicious behaviors related to order-swiping. [00...

AI Technical Summary

Problems solved by technology

[0005] The detection method in the prior art is limited to the self-inspection mechanism within the e-commerce platform. For the shops that are suspected to be fraudulent in the detection results of the fraudulent billing, the major e-commerce platforms only lower or force some of them. Closed and not disclosed to the public, users cannot obtain an intuitive data to view the status of order swiping. Based on the principle of maximizing benefits and the current development status of e-commerce, e-commerce platforms cannot fundamentally eliminate the phenomenon of order swiping

Images