汉语词的极性：基于新加坡华语使用者的极性词库建设 = Polarity of Chinese words : the construction of polarity Lexicon database based on Singapore Chinese speakers

• Main Author: 黄俊龙 Ng, Chin Loong
• Other Authors: Lin Jingxia
• Format: Final Year Project
• Language: Chinese
• Published: 2015
• Subjects: DRNTU::Humanities::Language::Chinese
• Online Access: http://hdl.handle.net/10356/65524
• Institution: Nanyang Technological University

情感分析研究领域众多，例如情感分类、情感强度、意见挖掘等，本文以研究汉语词汇的情感极性为主。纵观前人研究，情感极性主要采用计算机算法抑或通过寥寥数位研究人员来为词汇进行极性标注。然而，少数几位研究人员的情感标注其实无法代表普罗大众的看法。因此，本研究采集多位非专业人士（即普通的汉语使用者）对汉语词语的极性判断，并对此进行数据统计获得汉语词汇的极性均值。此外，本研究还讨论了情感极性与情感类别及情感强度之间的相关性，并与前人的极性研究结果进行了比较分析。 由于研究时间和资源有限，本研究的调查对象仅限于新加坡华语使用者。然而，本研究的调查结果所呈现的是首个基于大量普通的华语使用者的判断而建立的汉语情感词极性词库，本文所呈现的研究方法和结果对将来更为大型的汉语情感词极性研究或相关的情感研究都具有重大参考意义与借鉴作用。There are many research papers on sentiment analysis, however, those research mainly acquire the machine learning methods and rely heavily on the judgement of a few experts in obtaining the polarity of the words. Since the judgement of a few experts can hardly represent the judgement of the majority, this paper is the first to report a large-scale online survey of average language users’ judgement of perceived polarity scales on Chinese emotion words. Furthermore, this study also compared the current results of sentiment polarity with previous papers on emotion type, emotion intensity and sentiment polarity. As a pilot study, the results and methodology of this thesis will serve as an important reference for future research on sentiment analysis and emotion studies.