Advanced Sentiment Classification of Tibetan Microblogs on Smart Campuses Based on Multi-Feature Fusion

Sentiment analysis is an important problem in natural language processing, which plays an important role in many fields, such as information forecasting, knowledge classification, and product review. Because Tibetan microblogs have their own unique form, particularly the heterogeneous features, such as the emoticons, the grammatical relations, and the speech, the existing sentiment analysis method has difficulty in analyzing the emotions that such microblogs express. In this paper, we propose a sentiment classification method for Tibetan microblogs based on multi-feature fusion. To better study the affection of affective features, this paper first determines the theme of Weibo texts and chooses smart campuses as theme of Weibo texts for analyzing the influence of each feature on the sentiment of the microblog. Then, these features are fused as a multi-feature, and the sentiment of the Tibetan microblog is classified according to the multifeature fusion. The experimental results demonstrated that the sentiment classification algorithm based on feature fusion improved the accuracy of microblog sentiment classification.