A Recommendation System Based on Text Mining
Dongchen Miao, Fei Lang · 2017
With the improvement of social consumption level and the rapid development of the Internet, the application of recommendation system is becoming more and more extensive. Du to the complexity of Chinese language, the traditional recommendation system cannot grasp the user's sentiment tendencies well. In this paper, we establish a recommendation system with text mining technology. The proposed system uses the improved logistic regression in sentiment analysis to get user's sentiment score. Moreover, we build an item-feature matrix to calculate the feature similarity of the items, enhanceing the accuracy of item similarity. The experimental results demonstrate the effectiveness of our proposed system.