Emotion and Cause Linguistic Theory and Computational Implementation

This work argues that cause events, being the most tangible component of emotion, provide a rich dimension of how emotions should be classified. While it is often claimed that emotional concepts cannot be defined, this work views emotion as a response triggered by actual or perceived events, specifically focusing on the interaction between five primary emotions (Happiness, Sadness, Fear, Anger, and Surprise) and cause events. Cause events are examined in terms of two dimensions, namely transitivity and epistemicity. By incorporating the semantic and syntactic information of emotion cause events, this representation of emotion not only provides deep linguistic criteria of emotion cause events, but also offers an event-based approach to emotion classification. A text-driven, rule-based system for detecting the causes of emotion is then developed to establish the validity of the proposed linguistic model for emotion detection and classification. The system shows promising results.