SINAI: Voting System for Aspect Based Sentiment Analysis

Proceedings of the 8th International Workshop on Semantic Evaluation (SemEval 2014), pages 566–571, Dublin, Ireland, August 23-24, 2014. SINAI: Voting System for Aspect Based Sentiment Analysis Salud Mar´ıa Jim´enez-Zafra, Eugenio Mart´ınez-C´amara, M. Teresa Mart´ın-Valdivia, L. Alfonso Ure˜na-L´opez SINAI Research Group University of Ja´en E-23071, Ja´en (Spain) {sjzafra, emcamara, maite, laurena}@ujaen.es Abstract This paper describes the participation of the SINAI research group in Task 4 of the 2014 edition of the International Work- shop SemEval. This task is concerned with Aspect Based Sentiment Analysis and its goal is to identify the aspects of given target entities and the sentiment ex- pressed towards each aspect. 1 Introduction The web has evolved progressively since its be- ginning in 1990. At first, the user was almost a passive subject who received the information or published it, without many possibilities to gener- ate an interaction. The emergence of the Web 2.0 was a social revolution, because it offered users the possibility of producing and sharing contents, opinions, experiences, etc. Some years ago it was common to ask family and friends to know their opinion about a particu- lar topic, but after the emergence of the Web 2.0, the number of Internet users has been greatly in- creased. The exponential growth of the subjective information in the last years has created a great in- terest in the treatment of this information. Opinion Mining (OM), also known as Senti- ment Analysis (SA) is the discipline that focuses on the computational treatment of opinion, sen- timent and subjectivity in texts (Pang and Lee, 2008). Currently, OM is a trendy task in the field of Natural Language Processing due mainly to the fact of the growing interest in the knowledge of the opinion of people from different sectors of the society. However, the study on Opinion Mining goes back to 2002 when two of the most cited arti- This work is licensed under a Creative Commons At- tribution 4.0 International Licence. Page numbers and pro- ceedings footer are added by the organisers. Licence details: http://creativecommons.org/licenses/by/4.0/ cles in this task were published (Pang et al., 2002) (Turney, 2002). OM or SA can be divided into two subtasks that are known as subjectivity classification and polarity classification. Subjectivity classification is the task concentrated on the identification of subjectivity in texts, that is, these systems are bi- nary classifiers that separate the documents in two classes, objective and subjective ones. On the other hand, polarity classification is the task of de- termining the semantic orientation of a subjective text. The ideal OM system has to be composed by a subjectivity classifier and a polarity classifier. However, most of the works in the field of OM are carried out considering the documents as subjec- tive, so polarity classification systems have been more studied than subjectivity classification ones. The reader can find a complete overview about the research in OM in (Pang and Lee, 2008) and (Liu, 2012). As Liu asserts in (Liu, 2012), the polarity clas- sification systems can be divided into three …

Content truncated to 500 words