Empirical Study on Credit Classification of E - commerce Sellers Based on FCM Algorithm

Abstract

E-commerce platform evaluates e-seller star level according to the number of favourable comments from e-buyers. However, there is often inconsistence between the e-buyer evaluation details and the evaluation result. The evaluation result includes subjective factors in some degrees, so the e-seller star level which is only determined by the amount of favourable comments cannot completely reflect the e-seller credit level. Therefore, the shoes e-seller on taobao e-commerce platform will be taken as an example to evaluate the e-seller credit. Firstly, nine evaluation indexes such as the product quality, the product description, the service, the sales volume and so on, which embraced in evaluation details are selected. One hundred and fifty-three shoes e-seller samples are extracted by means of Python. It obtains shoes e-seller classification using fuzzy c-means clustering algorithm, and we devise their credit rank according to classification result. E-seller credit rank is compared with their star level, and their difference is analyzed. Secondly, we use FCM algorithm to analyze the relevance between each index and credit, so as to determine the main indicators affecting the e-seller credit. The evaluation mechanism of the e-seller credit designed above can provide the reference for e-buyer to make decision.