A hybrid method for sentiment classification in Chinese Movie Reviews based on sentiment labels

摘要

This paper proposes a method of sentiment classification in Chinese movie reviews. Given a Chinese Movie Review, we classify its sentiment as positive, neutral or negative. We introduce a hybrid approach based on sentiment labels which combining semantic based method with Support Vector Machine(SVM) machine learning method. We regard each review text as a sentiment phrases sequence, the approach first obtains two potential sentiment labels for each Chinese movie review by using semantic based and machine learning based approach respectively, and then regards hybrid labels as new features for machine learning based classification to improve its performance. Experimental results using hybrid method based on sentiment labels show the effectiveness and superior performance of the proposed approach.