Applying uncertainty theory into the restaurant recommender system based on sentiment analysis of online Chinese reviews

In this study, we utilize users' reviews to a restaurant recommender system to further explore users' opinions by the proposed recommender approach. Considering the uncertainty of users opinions, we apply the uncertain set to acquire users' sentiment polarity, and the uncertain variable to determine users' sentiment strength through sentiment analysis. To more accurately identify users' opinions, a distance-based approach is designed to detect the similar reviewers' opinions by combining sentiment polarity and sentiment strength. And then, a restaurant recommender model is proposed to evaluate the effectiveness of the presented recommendation algorithm. In experiments, we tested the performance of the recommendation algorithm with two real-world data sets. Even more remarkable, we compared the proposed user's profile that are used in two experiments. The experiments demonstrate that the significant performance of our method in terms of increasing the accuracy of the recommender system results. These results show that user-provided reviews include richer information than ratings in the process of recommender. We also uncover the effectiveness of the uncertainty theory to characterize users' opinions.