A CCG-based Approach to Fine-Grained Sentiment Analysis

摘要: In this paper, we present a Combinatory Categorial Grammar (CCG) based approach to the classification of emotion in short texts. We develop method that makes use notion put forward by Ortony et al. (1988), emotions are valenced reactions. This hypothesis sits central our system, which adapt contextual valence shifters infer emotional content text. integrate with an augmented version WordNet-Affect, acts as lexicon. Finally, experiment corpus headlines proposed 2007 SemEval Affective Task (Strapparava and Mihalcea, 2007), taking other competing systems baseline, demonstrate categorisation performs favourably.