Automated Narrative Information Extraction Using Non-Linear Pipelines / 4036
Our research focuses on the problem of automatically acquiring structured narrative information from natural language. We have focused on character extraction and narrative role identification from a corpus of Slavic folktales. To address natural language processing (NLP) issues in this particular domain we have explored alternatives to linear pipelined architectures for information extraction, specifically the idea of feedback loops that allow feeding information produced by later modules of the pipeline back to earlier modules. We propose the use of domain knowledge to improve core NLP tasks and the overall performance of our system.