Construction of a Blog Emotion Corpus for Chinese Emotional Expression Analysis

Proceedings of the 2009 Conference on Empirical Methods in Natural Language Processing, pages 1446–1454, Singapore, 6-7 August 2009. c⃝2009 ACL and AFNLP Construction of a Blog Emotion Corpus for Chinese Emotional Expression Analysis Changqin Quan Faculty of Engineering University of Tokushima 2-1 Minamijosanjima Tokushima Japan quan-c@is.tokushima-u.ac.jp Fuji Ren Faculty of Engineering University of Tokushima 2-1 Minamijosanjima Tokushima Japan ren@is.tokushima-u.ac.jp Abstract There is plenty of evidence that emotion analysis has many valuable applications. In this study a blog emotion corpus is con- structed for Chinese emotional expression analysis. This corpus contains manual an- notation of eight emotional categories (ex- pect, joy, love, surprise, anxiety, sorrow, angry and hate), emotion intensity, emo- tion holder/target, emotional word/phrase, degree word, negative word, conjunction, rhetoric, punctuation and other linguistic expressions that indicate emotion. An- notation agreement analyses for emotion classes and emotional words and phrases are described. Then, using this corpus, we explore emotion expressions in Chi- nese and present the analyses on them. 1 Introduction Textual emotion analysis is becoming increasingly important due to augmented communication via computer mediated communication (CMC) inter- net sources such as weblogs, email, websites, fo- rums, and chat rooms. Especially, blogspace con- sists of millions of users who maintain an online diary, containing frequently-updated views and personal remarks about a range of issues. Despite the increased focus on analysis of web content, there has been limited emotion analy- sis of web contents, with the majority of studies focusing on sentiment analysis or opinion min- ing. Classifying the mood of a single text is a hard task; state-of-the-art methods in text classi- fication achieve only modest performance in this domain (Mishne, 2005). In this area, some of the hardest problems involve acquiring basic re- sources. Corpora are fundamental both for devel- oping sound conceptual analyses and for training these emotion-oriented systems at different lev- els: to recognize emotions, to express appropriate emotions, to anticipate emotions, and other emo- tion processing applications. In this study we propose a relatively fine- grained annotation scheme, annotating emotion in text at three levels: document, paragraph, and sen- tence. We select eight emotion classes (expect, joy, love, surprise, anxiety, sorrow, angry and hate) for this annotation, and explore various linguis- tic expressions that indicate emotion in Chinese. The annotation scheme has been employed in the manual annotation of a corpus containing 1,487 documents, with 11,255 paragraphs, 35,096 sen- tences, and 878,164 Chinese words. Then, using this corpus, we explore and present data analy- ses on emotions, involving emotion states, accom- panying emotions, transfer emotions, independent emotions in texts. The remainder of this paper is organized as fol- lows. Section 2 describes the emotion corpus an- notation scheme. Section 3 presents the inter- annotator agreement study. Section 4 describes the analysis of emotion expressions. Section 5 presents a review of current emotion corpora for textual emotion analysis. Section 6 concludes this study with closing remarks and future directions. 2 Blog Emotion Corpus Annotation Scheme Weblogs are an increasingly popular mode of communication in the ever changing …

Content truncated to 500 words