Our system, AESOP, automatically creates plot unit representations for narrative text. AESOP has four main steps: affect state recognition, character identiﬁcation, affect state projection, and link creation. During affect state recognition, AESOP identiﬁes words that may be associated with positive, negative, and mental states. AESOP then identiﬁes the main characters in the story and applies affect projection rules to map the affect states onto these characters. During this process, some additional affect states are inferred based on verb argument structure. Finally, AESOP creates cross-character links and causal links between affect states. We also present two corpus-based methods to automatically produce a new resource for affect state recognition: a patient polarity verb lexicon.
It was developed by Amit Goyal and Ellen Riloff and Hal Daume. First, we collected a large collection of AESOP fables which are available here. Second, development, tune, and test set fables are available here. Third, the plot unit annotations for test set are available here. Fourth, the PPV lexicons are available here.
If you use any of the resources on this page, please include a citation to one of the publications below:References: