Research map / Label-efficient and robust learning / Recognition and long-tail
Corpus Wide Argument Mining—A Working Solution
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 |
|---|---|---|---|
| Train BA MaskS on VLD instead of the prior training data | BA MaskS · BlendNet | accuracy: 0.74 -> 0.78 | Helped |
| Add motion text to BA MaskS input (ablation) | BA MaskS · BlendNet | accuracy: 0.78 -> 0.77 | Hurt |
| Replace BA with BERT while using masked sentence input (ablation) | sentence evidence classifier · BlendNet | accuracy: 0.78 -> 0.81 | Helped |
| Add motion text to BERT MaskS input (ablation) | BERT MaskS · BlendNet | accuracy: 0.81 -> 0.82 | Helped |
| Use unmasked sentence text instead of masking the topic (ablation) | BERT S+M · BlendNet | accuracy: 0.82 -> 0.84 | Helped |
| Add motion text to classifier input (ablation) | BA and BERT evidence classifiers · VLD end-to-end retrieval | precision | Helped |
| Mask the topic in the sentence (ablation) | BA evidence classifier · VLD end-to-end retrieval | precision | Helped |
| Mask the topic in the sentence (ablation) | BERT evidence classifier · VLD end-to-end retrieval | precision | Hurt |
| Lower the classification threshold from 0.5 to 0.002 | BERT S+M · UKP-TUDA | recall: 0.16 -> 0.75 | Helped |
| Lower the classification threshold from 0.5 to 0.002 | BERT S+M · UKP-TUDA | precision: 0.88 -> 0.66 | Hurt |
Closest papers by meaning
- OpenDebateEvidence: A Massive-Scale Argument Mining and Summarization Dataset (NeurIPS 2024 · similarity 0.90)
- Argument Mining from Speech: Detecting Claims in Political Debates (AAAI 2016 · similarity 0.88)
- ARQUSUMM: Argument-aware Quantitative Summarization of Online Conversations (AAAI 2026 · similarity 0.87)
- A Large-Scale Dataset for Argument Quality Ranking: Construction and Analysis (AAAI 2020 · similarity 0.87)
- Fine-Grained Argument Unit Recognition and Classification (AAAI 2020 · similarity 0.87)
- Argument Mining for Improving the Automated Scoring of Persuasive Essays (AAAI 2018 · similarity 0.87)
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.