情绪原因识别是情绪分析中的一个重要研究任务。该任务旨在自动分析出导致某一情绪发生的原因描述。该文将情绪原因识别任务建模为序列标注模型,即将情绪词相关的子句当成序列,进而整体标注出哪些属于原因子句。具体实现中,我们使用条件随机场(CRF)模型进行求解,并结合了基本词特征、词性特征、距离特征、上下文特征及语言学特征等多种特征进行原因识别。实验结果表明,所采用的这些特征对于原因识别都有一定帮助,特别是上下文特征。此外,我们发现在使用类似特征集合的情况下,序列标注模型能够获得比分类模型更好的识别效果。
Abstract
Emotion cause detection is an important task in the research on emotion analysis. This task aims to detect the cause description of a emotion happening. In this study, we model this task as a sequence labeling problem and predict each related sentence to be in a emotion cause or not. Specifically, we apply the conditional random field (CRF) model to solve this problem with various of features, such as basic word features, POS features, context features and linguistic rule features. Empirical studies demonstrate that these features are effective for the task, especially the context features. Moreover, we find that the sequence labeling model is superior to the classification model when similar features are employed.
Key wordssequence labeling; emotion cause detection; context feature; linguistic rule features
关键词
序列标注 /
情绪原因识别 /
上下文特征 /
语言学规则特征
{{custom_keyword}} /
Key words
sequence labeling /
emotion cause detection /
context feature /
linguistic rule features
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] Alm C, D Roth, R Sproat. Emotions from Text: Machine Learning for Text-based Emotion Prediction[C]//Proceedings of EMNLP-05, 2005: 579-586.
[2] Mihalcea R, H Liu. A Corpus-based Approach to Finding Happiness[C]//Proceedings of the AAAI Spring Symposium on Computational Approaches to Weblogs. 2006.
[3] Tokuhisa R, K Inui Y Matsumoto. Emotion Classification Using Massive Examples Extracted from the Web[C]//Proceedings of COLING. 2008: 881-888.
[4] Descartes R. 1649. The Passions of the Soul[M]. J. Cottingham et al. (Eds), The Philosophical Writings of Descartes. 2008, Vol(1): 325-404.
[5] James W. 1884. What is an Emotion? Mind,9(34):188-205.
[6] Plutchik R. Emotions: A Psychoevolutionary Synthesis[M]. New York: Harper & Row.1980.
[7] Wierzbicka A. Emotions across Languages and Cultures: Diversity and Universals[M]. Cambridge: Cambridge University Press. 1999.
[8] Lee S, Chen Y, Huang C et al. Detecting Emotion Causes with a Linguistic Rule-based Approach[J]. Computational Intelligence. 2012.
[9] Chen Y, S Lee, S Li, et al. Emotion Cause Detection with Linguistic Constructions[C]//Proceeding of COLING-10. 2010: 179-187.
[10] Pang B, L Lee. Opinion Mining and Sentiment Analysis: Foundations and Trends[J]. Information Retrieval, 2008, vol.2(12): 1-135.
[11] Cui H, V Mittal, M Datar. Comparative Experiments on Sentiment Classification for Online Product Comments[C]//Proceedings of AAAI-06. 2006, 1265-1270.
[12] Li S, C Huang, G Zhou, et al. Employing Personal/Impersonal Views in Supervised and Semi-supervised Sentiment Classification[C]//Proceedings of ACL-10. 2010, 414-423.
[13] Kim S, E Hovy. Identifying Opinion Holders for Question Answering in Opinion Texts[C]//Proceedings of the Workshop on Question Answering in Restricted Domain at AAAI-05.2005.
[14] Xu G, X Meng, H Wang. Build Chinese Emotion Lexicons Using A Graph-based Algorithm and Multiple Resources[C]//Proceeding of COLING-10. 2010: 1209-1217.
[15] Volkova S, W Dolan, T Wilson. CLex: A Lexicon for Exploring Color, Concept and Emotion Associations in Language[C]//Proceedings of EACL-12. 2012: 306-314.
[16] Quan C, F Ren. Construction of a Blog Emotion Corpus for Chinese Emotional Expression Analysis[C]//Proceedings of EMNLP-09. 2009: 1446-1454.
[17] Purver M, S Battersby. Experimenting with Distant Supervision for Emotion Classification[C]//Proceeding of EACL-12. 2012: 482-491.
[18] Lafferty J, McCallum A, Pereira F. Conditional Random Fields: Probabilistic Models for Segmenting and Labeling Sequence Data[C]//Proceedings of ICML-2001. 2001: 282-289.
[19] 宗成庆. 统计自然语言处理[M]. 清华大学出版社: 北京,2008.
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
基金
香港GRF项目(543810),国家自然科学基金资助项目(61003155,61273320)
{{custom_fund}}
PDF(1071 KB)
