作者: Cynthia Van Hee , Els Lefever , Veronique Hoste
DOI: 10.18653/V1/S15-2115
关键词:
摘要: This paper describes our contribution to the SemEval-2015 Task 11 on sentiment analysis of figurative language in Twitter. We considered two approaches, classification and regression, provide fine-grained scores for a set tweets that are rich sarcasm, irony metaphor. To this end, we combined variety standard lexical syntactic features with specific capturing content. All experiments were done using supervised learning LIBSVM. For both runs, system ranked fourth among fifteen submissions.