초록 열기/닫기 버튼
Eunkyoung Jo. 2015. Extracting Lexical functional words and Aboutness in Sentiment Analysis. Korean Semantics, 47. These days, there are enormous volumes of documents which are written by internet users. And we would like to figure out what aboutness those documents have in the aspects of sentiment and to find out some lexical functional words which don't have semantic polarity nor aboutness but play a role in determining the polarity. For doing so, we utilize statistical methods of word weighting scheme named TF-IDF(Term Frequency-Inversed Document Frequency) for extracting lexical functional words and of topic modeling named 'LDA(Latent Dirichlet Allocation)' for extracting aboutness. Overall, this paper deals with extracting lexical functional word and topic words of aboutness that will be used for improving the result of sentiment analysis. Therefore, we examined the effects of the lexical functional words as quantitative evaluation and reviewed the effects of including topic words as qualitative evaluation. Furthermore, we are sure that statistical analytic methods used in this paper would be useful for text sentiment analysis.
키워드열기/닫기 버튼
sentiment analysis, lexical functional word, aboutness, IDF:inversed document frequency, LDA:latent dirichlet allocation, topic word, movie corpus, statistical approach
