Analysis of C2C Website Competitiveness Based on Text Sentiment Analysis and Dynamic Fuzzy Computation

With the global step into the era of large data, online review data show an explosive growth trend, and the resulting information overload has affected the consumer’s decision making. In terms of consumers, to tap the purchase decision-making information from the vast amount of information is a must-do before the tedious process; how to find the best service from a variety of shopping sites and find satisfied products from thousands of Products and how to help companies improve themselves, has become the forefront of e-commerce research topics. In this paper, data acquisition, data filtering, theme analysis, correlation analysis, sentiment analysis and dynamic fuzzy statistics are introduced to the user-generated content (UGC) fuzzy comprehensive evaluation. Firstly, collect the data, we use the correlation analysis and the stopped words to reduce the dimensionality of the data, and generate the revised candidate sets of themes. Then, we use emotion extraction, emotion polarity judgment and comment analysis to divide the UGC into several evaluation grades; then we get the fuzzy comprehensive evaluation matrix by the fuzzy statistic method. Finally, the comprehensive evaluation model is constructed to give users product score and the C2C website ranking.