Research map / Label-efficient and robust learning / Recognition and long-tail
Implicit Discourse Relation Classification via Multi-Task Neural Networks
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 |
|---|---|---|---|
| multi-task learning with auxiliary discourse corpora (ablation) | CNN implicit relation classifier · PDTB implicit 4-way | macro F1: 37.65 -> 44.98 | Helped |
| multi-task learning (ablation) | CNN classifier · PDTB implicit 4-way | accuracy: 52.82 -> 57.27 | Helped |
| add explicit PDTB relation task (Task 2) (ablation) | main task · PDTB Comparison | F1: 32.35 -> 25.88 | Hurt |
| add explicit PDTB relation task (Task 2) (ablation) | main task · PDTB Temporal | F1: 16.67 -> 27.27 | Helped |
| add RST-DT relation task (Task 3) (ablation) | main task · PDTB Comparison | F1: 32.35 -> 34.65 | Helped |
| add connective word classification task (Task 4) (ablation) | main task · PDTB Contingency | F1: 35.35 -> 44.27 | Helped |
| add connective word classification task (Task 4) (ablation) | main task · PDTB Comparison | F1: 32.35 -> 22.01 | Hurt |
| all tasks combined (ablation) | main task · PDTB Comparison | F1: 32.35 -> 32.13 | No clear effect |
| all tasks combined (ablation) | main task · PDTB Expansion | F1: 66.23 -> 69.88 | Helped |
| MTL vs R&X 2015 | implicit relation · PDTB binary Comparison | F1: 41.00 -> 37.91 | Hurt |
| MTL vs R&X 2015 | implicit relation · PDTB binary Temporal | F1: 33.30 -> 37.17 | Helped |
Closest papers by meaning
- Linguistic Properties Matter for Implicit Discourse Relation Recognition: Combining Semantic Interaction, Topic Continuity and Attribution (AAAI 2018 · similarity 0.92)
- Working Memory-Driven Neural Networks with a Novel Knowledge Enhancement Paradigm for Implicit Discourse Relation Recognition (AAAI 2020 · similarity 0.89)
- A Label Dependence-Aware Sequence Generation Model for Multi-Level Implicit Discourse Relation Recognition (AAAI 2022 · similarity 0.88)
- Improving Implicit Discourse Relation Recognition with Natural Language Explanations from LLMs (AAAI 2026 · similarity 0.87)
- A Deep Sequential Model for Discourse Parsing on Multi-Party Dialogues (AAAI 2019 · similarity 0.86)
- ImpScore: A Learnable Metric For Quantifying The Implicitness Level of Sentences (ICLR 2025 · similarity 0.80)
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.