Analysis And Detection Of False Trading Commodities Based On Distributed Clustering In Big Data

The progress and development of industry is always closely related to the progress of technology.With the rise of the Internet industry,online trading platform has gradually entered our lives.However,in the process of online trading,due to the virtual attributes of the network itself,there are false transactions.Businessmen use the form of employment to deal falsely with a certain commodity in order to obtain the right to evaluate the commodity through the way of falsely trading,and then drive the trend of commodity comment on the whole.Because online transactions do not actually have access to commodities,they will refer to the comments.Businessmen achieve the purpose of influencing consumers' decision-making through the form of false transactions.With the industrialization of false transactions,these behaviors become organizational.In order to identify the false transactions of commodities,we usually start from reviews,and find this kind of organizational account group by reviewing text,reviewing behavior and other characteristics.However,these works are not aimed at the false trading commodities themselves,there are some limitations.This paper presents a method of analysis and detection of false trading commodities in large data scenarios.By extracting transaction records of accounts,calculating similarity between multiple accounts and clustering,we can get false trading account groups.Then,we use four-fold group transaction fraud index to score candidate groups,select false trading groups,and finally calculate the false trading groups in commodity transaction records.The proportion of the goods,the suspicion of the goods.At the same time,this paper designs and implements a distributed system that meets the new generation of microservice standard,Service Mesh,based on the above algorithm,considering the application in big data scenario and referring to the industry's microservice architecture and Sidecar design pattern.The system has goodscalability and maintainability,realizes the software design idea of separation of concerns,and realizes the algorithm of false trading commodities proposed in this paper.