Institute of Formal and Applied Linguistics

Coreference resolution research has long been dominated by English-centric evaluation, partly due to the heterogeneity of available resources. To address this imbalance, my colleagues and I have been developing CorefUD, a growing multilingual collection of datasets annotated with coreference and anaphoric relations. We also organize the CRAC Shared Tasks, aiming to promote cross-linguistic research in this area.

In this talk, I will share our progress over the past four years, the challenges we have faced, and discuss to what extent our efforts have driven meaningful change in the field. Finally, I will reflect on the role of coreference-annotated resources and dedicated resolution systems in the era of large language models.

*** The talk will be delivered in person (MFF UK, Malostranské nám. 25, 4th floor, room S1) and will be streamed via Zoom. For details how to join the Zoom meeting, please write to sevcikova et ufal.mff.cuni.cz ***