Text Sentiment Analysis Based on Amazon Product Evaluation

This article expounds and demonstrates how to help the Amazon online products gain the favor of the users. It selects the user evaluation text from some Amazon products. It uses the methods of data analysis, coefficient of variation, literature review and questionnaire to select the relevant indicators. Nine factors are chosen as indicators such as average product score, total number of comments and number of helpful comments and so on. The user preference evaluation index system is established, and the weight is determined by swing weight method. Based on the theory and method of grey fuzzy comprehensive evaluation and text emotion analysis, a comprehensive evaluation model is established, and the evaluation results and suggestions are obtained.