Research on Reader Sentiment Analysis Model Research Based on Online Book Reviews
ZHANG Ji;School of Computer Science and Engineering, Anhui University Of Science & Technology;
User reviews in e-book website truly reflect reader feeling about e-book,so the research of sentiment analysis of book re-views has academic and commercial value. Discussing on machine learing models for sentiment analysis,using sentiment lexicon assubjective content recognition,through the neural network language model to extend the emotional dictionary, the use of machinelearning methods, design evaluation of the field of emotional analysis model.Analysis of the logic regression classification algo-rithm, the use of a variety of evaluation indicators. analysis of different model results, the results show that the use of subjective andobjective classification of the model is more suitable to judge the reader's emotional tendencies.
【CateGory Index】： TP18;TP391.1