Narrative is one of the most common forms of discourse and it is all around us, helping us make sense of our everyday world. Narratives can move from one form to another, from a short story to a novel to a movie, or even be retold in the same medium, but we can still distinguish them as having the same story. The question then arises of how this similarity can be defined and how it can be detected automatically. I work from a hypothesis that two narratives can be said to be similar if their important (kernel) events are similar. Extensive work has been done in the field of information extraction relating to event detection and extraction, but more effort has yet to be devoted to kernel event detection and extraction. My proposed dissertation project has three main objectives. First, I aim to improve the current baseline kernel event detection and extraction system that I have been working on by incorporating document-level context into the model. Second, I propose to add the concept of causal relationship that exists between the kernel events in a narrative to further expand the capability of the system. Third, I plan to explore the similarity between different narratives by comparing the kernel events extracted from narratives. In addition to the direct deliverables, this research project will enable deeper narrative understanding in automated systems and aid in tasks such as summarization, narrative generation, and question-answering. It will also provide a framework for further work on automating mis/disinformation detection by identifying similar narratives using kernel events.

Anshu Kiran Sharma is a Ph.D. candidate at the Knight Foundation School of Computing & Information Sciences (KFSCIS) at Florida International University (FIU), under the supervision of Professor Mark A. Finlayson. She is a part of the Cognac laboratory (Cognition, Narrative, and Culture). Her research interests include Natural Language Processing (NLP), Computational Linguistics, Narratives, and Generative AI. Anshu holds a B.S. degree in Computer Science from Tribhuvan University, Kathmandu, Nepal.