A Hybrid Method with Text Mining and Multi-criteria Decision Making for E-Commerce Considering Online Reviews

Abstract

AbstractThis study intends to hybrid three methods for online reviews analysis through text mining, Latent Dirichlet Allocation (LDA), fuzzy decision making trial and evaluation laboratory (FDEMATEL), and fuzzy technique for order of preference by similarity to ideal solution (FTOPSIS) to find the causal relationship between complex e-commerce factors affecting customers decision making and in selecting the best e-commerce sites. The original idea of the proposed method is established in the context of e-customers satisfaction review, especially for customer segmentation in fashion and beauty shopping through customers’ online reviews. We use text mining and LDA for analyzing textual reviews, FDEMATEL to reveal the causal relationships on customers do online shopping, and FTOPSIS for ranking the best e-commerce platform. The impact of causal relationships on e-commerce services is investigated for six different criteria. The proposed method is appraised based on the customers’ reviews on 4 e-commerce sites in Malaysia. The finding of this study provides the best choice method for customers’ decision-making for e-commerce site selection and helps e-commerce managers to increase profit from the service quality, marketing tactics, and strategies.KeywordsText miningLDAMultiple criteria decision-makingFDEMATELFTOPSISOnline reviewE-commerce