Abstract
We evaluate a rule-based (Lee et al., 2013) and neural (Lee et al., 2018) coreference system on Dutch datasets of two domains: literary novels and news/Wikipedia text. The results provide insight into the relative strengths of data-driven and knowledge-driven systems, as well as the influence of domain, document length, and annotation schemes. The neural system performs best on news/Wikipedia text, while the rule-based system performs best on literature. The neural system shows weaknesses with limited training data and long documents, while the rule-based system is affected by annotation differences. The code and models used in this paper are available at https://github.com/andreasvc/crac2020
| Original language | English |
|---|---|
| Title of host publication | Proceedings of the Third Workshop on Computational Models of Reference, Anaphora and Coreference |
| Editors | Maciej Ogrodniczuk, Vincent Ng, Yulia Grishina, Sameer Pradhan |
| Publisher | Association for Computational Linguistics (ACL) |
| Pages | 79-90 |
| Number of pages | 12 |
| Publication status | Published - 2020 |
| Event | Workshop on Computational Models of Reference, Anaphora and Coreference - Online Duration: 12-Dec-2020 → … Conference number: 3 |
Workshop
| Workshop | Workshop on Computational Models of Reference, Anaphora and Coreference |
|---|---|
| Period | 12/12/2020 → … |