Sentiment Analysis of Music Criticism Based Data Mining

摘要

With the development of web2.0, more and more people like to express their emotions by making comments anytime and anywhere. Music comments are an emotional expression of the listener's mood for listening to songs. On the basis of the original Hevner emotion loop, this paper proposes an optimized emotional model that fits Chinese people's thinking and language habits, constructs a new musical sentiment dictionary, analyzes the polarity of musical emotions, and proposes an emotional vector space model based on emotional units. Through experiments, it is proved that the emotional vector space model has higher accuracy and convenience than artificial emotion annotation.