EmoBank: Studying the Impact of Annotation Perspective and Representation Format on Dimensional Emotion Analysis
This work addresses the challenge of emotion annotation in NLP by providing a balanced corpus with multiple perspectives and formats, which is incremental as it builds on existing emotion representation methods.
The authors tackled the problem of dimensional emotion analysis by creating EmoBank, a corpus of 10k English sentences annotated with Valence-Arousal-Dominance metadata, and found that the reader's perspective yields higher inter-annotator agreement and rating intensity, achieving close-to-human performance in mapping between dimensional and categorical formats.
We describe EmoBank, a corpus of 10k English sentences balancing multiple genres, which we annotated with dimensional emotion metadata in the Valence-Arousal-Dominance (VAD) representation format. EmoBank excels with a bi-perspectival and bi-representational design. On the one hand, we distinguish between writer's and reader's emotions, on the other hand, a subset of the corpus complements dimensional VAD annotations with categorical ones based on Basic Emotions. We find evidence for the supremacy of the reader's perspective in terms of IAA and rating intensity, and achieve close-to-human performance when mapping between dimensional and categorical formats.