Goods fleeing early warning method and device based on business full-chain big data analysis modeling

Through the early warning method of cross-stock goods based on business full-chain big data analysis and modeling, machine learning models are used to predict and warn the risk of cross-stock goods, which solves the problem that cross-stock identification and early warning algorithms in the existing technology cannot achieve real-time early warning and cover all sales scope, and realizes real-time early warning and effective stopping of cross-stock behavior.

CN119693035APending Publication Date: 2025-03-25INNER MONGOLIA YILI IND GROUP CO LTD
0 Cites 2 Cited by

Patent Information

Application Number
CN202311238382.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2023-09-22
Publication Date
2025-03-25

AI Technical Summary

Technical Problem

The existing intercept identification and early warning algorithms cannot achieve real-time early warning, and due to insufficient data types, lagging data timeliness or difficulty in obtaining data, these methods have great limitations and cannot cover the entire sales range.

Method used

The cross-trading early warning method based on business full-chain big data analysis and modeling is adopted, and data cleaning and feature extraction are cleaned by obtaining dealers' business data, and machine learning models (such as Bayesian Network and LightGBM) are used to predict and warn cross-trading risk.

Benefits of technology

Real-time early warning of goods cross-trading behavior is achieved, the audit cost is reduced, and the dealer's motives and abnormal situations are discovered in a timely manner, and effective stopping of goods cross-trading behavior has not yet occurred, thereby avoiding possible damage.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention discloses a commodity fleeing early warning method and device based on business full-chain big data analysis modeling, and the method comprises the steps: obtaining to-be-predicted operation data of a dealer, and obtaining first feature data; discretizing the first feature data, inputting the discretized first feature data into a constructed dealer channel conflict prediction model to obtain the probability of channel conflict of a dealer, extracting order data to be predicted from the cleaned operation data according to the probability of channel conflict of the dealer when the dealer has the risk of channel conflict, and predicting the order data to be predicted according to the extracted order data. Performing feature extraction on the order data to obtain second feature data; and inputting the second feature data into a constructed order goods fleeing prediction model to obtain the probability of goods fleeing of the order, and carrying out market goods fleeing early warning according to the probability of goods fleeing of the order. According to the method, the goods fleeing motivator and the abnormal condition of the dealer can be found in time, and the goods fleeing behavior at each stage is effectively stopped, so that possible damage is avoided.
Need to check novelty before this filing date? Find Prior Art