Sentiment Analysis of Film Review Texts Based on Sentiment Dictionary and SVM
Kui Lu, Jiesheng Wu · 2019
The sentiment analysis of the film review text is to extract and analyze the hidden sentiment information in the text data, thereby helping the network personnel such as the media platform to analyze the audience's preference for the film. Based on this, this paper proposes a film review text sentiment analysis method based on SVM classification technology in sentiment dictionary and maching learning. Firstly, the basic sentiment dictionary, the domin sentiment dictionary, the negative word dictionary and the degree adverb dictionary are constructed. The four dictionaries are combined to realize the expansion of the dictionary. Secondly, the SVM model training set is construcetd by calculating the combination of sentiment weight and user scoring. Finally, using the test data for sentiment classification experiments, the results show that the method has higher accuracy of sentiment classification than the method based on a basic sentiment dictionary.