Our approach exploits in an original way ‘crowd-sourced’ affective annotation implicitly provided by readers of news articles.
By using these massive crowd-sourced affective annotations over tens of thousands of news articles, and so called distributional semantics approaches, we were able to extract – in a totally automated way – a high-coverage and high-precision lexicon of roughly 37 thousand terms annotated with emotion scores, called DepecheMood.
DETAILS: all the technical details can be found in the paper: Jacopo Staiano, Marco Guerini, "DepecheMood: a Lexicon for Emotion Analysis from Crowd-Annotated News". To appear in Proceedings of ACL 2014. Preprint
RESOURCES: the Lexicon we developed - and that is used for the demo - is freely available for research purposes and it can be downloaded here.