Using MicroBlog public mood obtained from user-generated content data to predict company user feedback, has become a new approach in social internet study. In this paper, we conduct a forecast study on new products, based on MicroBlog sentiment analysis, a Chinese social network platform. First, we introduce a mood state theory and public mood word set, according to the Profile of Mood State (POMS), and obtain a semantic network of the mood words based on a synonym lexicon and Google word2vec. Then, we calculate the respective dimension values of the words and build a fine-grained lexicon of m...