Social Media Sentiment Analysis of Sequential Three-way Decision Model Based on Knowledge Fusion

Social media comment data has been one of the important research directions in online public opinion. To better deal with speech data of uncertain quality in a complex environment, this paper proposes a new model of text sentiment analysis based on knowledge fusion with sequential three-way decision making. The model incorporates external knowledge and sequential three-way decision theory by adding a sequential three-way decision step in local feature extraction based on the BERT-base model. Experiments show that significant performance improvements are achieved on all three social media datasets.