Identifying the Sentiment in Domain-Independent Chinese Sentences
This paper describes a system that can automatically extract the topics from subjective Chinese sentences and identify their emotional tendencies. In the topic identification phase, the system bases on the dictionary containing the emotional words, dependency grammar and statistical pattern to identify the topic. In the sentiment identification phase, the system bases on the modifying relationship between the adverbs, negation words and emotional words to identify the emotional tendency. Finally, we choose a large number of Chinese sentences as our benchmarks to evaluate the system’s performance, analyze its error rate and propose directions for our future improvements.
M. W. Si et al., "Identifying the Sentiment in Domain-Independent Chinese Sentences", Advanced Materials Research, Vols. 217-218, pp. 808-811, 2011