Automatic Judgment of the Subjectivity and Objectivity of the Chinese Words

The effective automatic judgment of the Chinese words sentiment polarity, the most important part of the Chinese sentiment analysis, can improve the building of the subjectivity lexicon and the efficiency of the sentiment analysis. The technology of the Chinese word subjectivity and objectivity judgment is discussed and analyzed, the subjectivity dictionary is defined and the subjective feature model is established by the use of the sentiment polarity of the word and the subjectivity intensity feature set. The machine learning method applied in the subjective feature set achieves the subjectivity classifier to automatically judge the word subjectivity and to compare and optimize. The performance is improved. The highest accuracy of KNN reaches 81.27%, and the F value is up to 81.52%. So it is effective to establish the word subjectivity feature set, using the sentiment polarity of and the subjectivity intensity feature the words. The automatic judgment of the word subjectivity through machine learning achieves excellent performance.