B2B Supplier’s Credibility Classification Based On Data Mining Technology

B2B e-commerce website provides a trading platform for purchasers and suppliers, as a third party transaction between the two roles. In order to provide better services for buyers, it is necessary to classify suppliers’reliability. As the B2B platform owns a large amount of data and is complex to express, this paper attempts to adopt data mining techniques to classify the credibility of the suppliers to create the classification model of credibility and provide valuable information for the purchaser enterprises.This paper is based on suppliers’information, exploits the statistical analysis and data mining technology to focus on suppliers sample data clustering and classification analysis.The paper aims at the K-Means algorithm and the ID3algorithm to study how they work and their advantages and disadvantages, and puts forward the improved algorithm Wm-K-Means algorithm and J-ID3algorithm which are achieved on the Eclipse development platform, and compares them with the traditional algorithms in the experiments. This paper uses the improved K-Means algorithm to cluster and analysis the sample data of the suppliers to form the suppliers’group in the clustering, marks the characteristic properties of each supplier group and uses the improved ID3algorithm to classify the credibility of suppliers to generate a decision tree and classification rules.At last,the credibility classification model of suppliers is formed.The paper describes the data mining process from the pre-processing of data to the cluster analysis and classification forecast of the data set in detail. The data mining results of the credibility classification is analyzed and the value and significance of the the supplier classification model obtained are discussed. These research work has some reference value for the actual B2B e-commerce site and also has a positive effect to improve the credibility of the suppliers.