Wait, did that actually happen? Graph-based event modality classification

Event extraction is a central NLP task which enables applications aiming to extract and aggregate information about real world events from texts. While there is considerable progress on neural modeling for event extraction there is still little work on the task of event modality classification, i.e. determining whether or not an event actually took place. This is a challenging task which may need to rely on linguistic cues from the larger context.

In this thesis, the existing EventGraph system (You et al, 2022, 2023) will be extended with capabilities for event modality classification and evaluations will be performed for English and Norwegian.

More information about the EventGraph system:

You, Huiling; Samuel, David; Touileb, Samia & Øvrelid, Lilja (2022). EventGraph: Event Extraction as Semantic Graph Parsing, The 5th Workshop on Challenges and Applications of Automated Extraction of Socio-political Events from Text. Association for Computational Linguistics. ISSN 978-1-959429-05-0. p. 7–15.
You, Huiling; Øvrelid, Lilja & Touileb, Samia (2023). JSEEGraph: Joint Structured Event Extraction as Graph Parsing. In Palmer, Alexis & Camacho-Collados, Jose (Ed.), Proceedings of the 12th Joint Conference on Lexical and Computational Semantics (*SEM 2023). Association for Computational Linguistics. ISSN 978-1-959429-76-0.

Publisert 3. okt. 2023 16:12 - Sist endret 3. okt. 2023 16:12

Veileder(e)

Huiling You Universitetet i Oslo
Lilja Øvrelid Universitetet i Oslo

Wait, did that actually happen? Graph-based event modality classification

Veileder(e)

Omfang (studiepoeng)