Researchers have introduced two new corpora, Hlava Cor and Hlava AD, designed to study human label variation in coreference and discourse relations. Hlava Cor contains 1,024 contexts annotated by three individuals, focusing on coreference identification across different linguistic elements. Hlava AD includes 512 contexts annotated by five individuals, concentrating on discourse relations. Both corpora exhibit an inter-annotator agreement of around 60-65%, with lower agreement observed in cases where automatic coreference resolution models also struggle, indicating ambiguity for human annotators. AI
IMPACT Highlights challenges in natural language understanding tasks, potentially guiding future model development for coreference and discourse.
RANK_REASON The cluster contains a research paper detailing new corpora for studying linguistic annotation variation.
- arXiv
- Czech
- Hlava Cor
- Hugging Face
- alphaXiv
- CatalyzeX
- CORE Recommender
- DagsHub
- Gotit.pub
- ScienceCast
AI-generated summary · Google Gemini · from 3 sources. How we write summaries →