Research map / Label-efficient and robust learning / Recognition and long-tail
Fine-Grained Argument Unit Recognition and Classification
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Add a CRF sequence-labeling layer (ablation) | BERT_LARGE in token-level AURC · AURC-8 in-domain test | token F1: .683 -> .696 | Helped |
| Add a CRF sequence-labeling layer (ablation) | BERT_LARGE in token-level AURC · AURC-8 cross-domain test | segment F1: .626 -> .649 | Helped |
| Add a CRF sequence-labeling layer (ablation) | BERT_LARGE in sentence-level AURC · AURC-8 in-domain test | sentence F1: .715 -> .725 | Helped |
| Add a CRF sequence-labeling layer (ablation) | BERT_LARGE in sentence-level AURC · AURC-8 cross-domain test | token F1: .544 -> .519 | Hurt |
| Use BERT_LARGE instead of BERT_BASE | token-level AURC · AURC-8 in-domain test | token F1: .654 -> .683 | Helped |
| Include the topic as an additional model input | BERT AURC models · AURC-8 | F1 | Helped |
| Annotate argument spans at token level rather than sentence level | AURC-8 argument retrieval annotations · AURC-8 total | number of argument units: +10.51 | Helped |
| Constrain spans with a slot-filling argument template | expert annotation of argument units · 160 sampled sentences | Krippendorff alpha_u_nom: 0.51 -> 0.61 | Helped |
| Use token-level rather than sentence-level setup without sentence-boundary information | BERT_LARGE models · AURC-8 cross-domain test, sequences crossing sentence boundaries | segment F1: .154 -> .538 | Helped |
Closest papers by meaning
- A Large-Scale Dataset for Argument Quality Ranking: Construction and Analysis (AAAI 2020 · similarity 0.87)
- Corpus Wide Argument Mining—A Working Solution (AAAI 2020 · similarity 0.87)
- OpenDebateEvidence: A Massive-Scale Argument Mining and Summarization Dataset (NeurIPS 2024 · similarity 0.86)
- Lawma: The Power of Specialization for Legal Annotation (ICLR 2025 · similarity 0.82)
- ACT as Human: Multimodal Large Language Model Data Annotation with Critical Thinking (NeurIPS 2025 · similarity 0.82)
- M3Grounder: Mask-Based Multi-Span and Multi-Granular Grounding for Document QA (CVPR 2026 · similarity 0.81)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.