OpAL: Applying Opinion Mining Techniques for the Disambiguation of Sentiment Ambiguous Adjectives in SemEval-2 Task 18

Proceedings of the 5th International Workshop on Semantic Evaluation, ACL 2010, pages 444–447, Uppsala, Sweden, 15-16 July 2010. c⃝2010 Association for Computational Linguistics OpAL: Applying Opinion Mining Techniques for the Disambiguation of Sentiment Ambiguous Adjectives in SemEval-2 Task 18 Alexandra Balahur University of Alicante Department of Software and Computing Systems abalahur@dlsi.ua.es Andrés Montoyo University of Alicante Department of Software and Computing Systems montoyo@dlsi.ua.es Abstract The task of extracting the opinion expressed in text is challenging due to different reasons. One of them is that the same word (in particu- lar, adjectives) can have different polarities depending on the context. This paper presents the experiments carried out by the OpAL team for the participation in the SemEval 2010 Task 18 – Disambiguation of Sentiment Ambiguous Adjectives. Our approach is based on three dif- ferent strategies: a) the evaluation of the polar- ity of the whole context using an opinion min- ing system; b) the assessment of the polarity of the local context, given by the combinations between the closest nouns and the adjective to be classified; c) rules aiming at refining the lo- cal semantics through the spotting of modifi- ers. The final decision for classification is tak- en according to the output of the majority of these three approaches. The method used yielded good results, the OpAL system run ranking fifth among 16 in micro accuracy and sixth in macro accuracy. 1 Credits This research has been supported by Ministerio de Ciencia e Innovación - Spanish Government (grant no. TIN2009-13391-C04-01), and Consel- leria d'Educación-Generalitat Valenciana (grant no. PROMETEO/2009/119 and ACOMP/2010/ 288). 2 Introduction Recent years have marked the beginning and ex- pansion of the Social Web, in which people free- ly express and respond to opinion on a whole variety of topics. Moreover, at the time of taking a decision, more and more people search for in- formation and opinions expressed on the Web on their matter of interest and base their final deci- sion on the information found (Pang and Lee, 2008). Nevertheless, the high quantity of data that has to be analysed imposed the development of specialized Natural Language Processing (NLP) systems that automatically extract, classi- fy and summarize the opinions available on the web on different topics. Research in this field, of opinion mining (sentiment analysis), has ad- dressed the problem of extracting and classifying opinions from different perspectives and at dif- ferent levels, depending on various factors. While determining the overall opinion on a mov- ie is sufficient for taking the decision to watch it or not, when buying a product, people are inter- ested in the individual opinions on the different product characteristics. Especially in this con- text, opinion mining systems are confronted with a difficult problem: the fact that the adjectives used to express opinion have different polarities depending on the characteristic they are men- tioned with. For example, “high price” is nega- tive, while “high resolution” is positive. There- fore, specialized methods have to be employed to correctly determine the contextual polarity of …

Content truncated to 500 words