Customer Satisfaction Evaluation Model of E-commerce Website based on Tensor Analysis

Abstract

Taking the e-commerce website as an example and combing with tensor analysis theory, a novel evaluation model of customer satisfaction on the website is established in higher order space. Firstly, the model takes use of tensor analysis to expand the index system of customer satisfaction with a high-order tensor. Then the customer satisfaction evaluation model is given by means of high order subspace analysis based on non-negative Tucker decomposition. Finally, the model training process and testing results are given with the data in real applications. Experimental results show that the model has feasibility and performs well.