《D09-1131（论文）》.pdf

Using Morphological and Syntactic Structures for Chinese Opinion Analysis Lun-Wei Ku Ting-Hao Huang Hsin-Hsi Chen Department of Computer Science and Information Engineering National Taiwan University No. 1, Sec. 4, Roosevelt Road, Taipei, 10617 Taiwan {lwku,thhuang}@.tw;hhchen@.tw composing characters. When the opinion words Abstract are available, the polarities of sentences and documents can be determined by them. Riloff This paper employs morphological struc- and Wiebe (2003) learned the extraction patterns tures and relations between sentence seg- for subjective expressions. Kim and Hovy (2004) ments for opinion analysis on words and found the polarity of subjective expressions. sentences. Chinese words are classified Pang et al. (2002) and Dave et al. (2003) ex- into eight morphological types by two plored various techniques at document level. proposed classifiers, CRF classifier and Morphological information has been widely SVM classifier. Experiments show that used in classifying words, telling the meanings, the injection of morphological information and doing other in-depth analysis (Tzeng and improves the performance of the word po- Chen, 2002). However, morphological informa- larity detection. To utilize sy