Research map / Architectures and efficiency / RNNs and normalization
Dependency or Span, End-to-End Uniform Semantic Role Labeling
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 |
|---|---|---|---|
| Use the unified end-to-end model for span SRL | Span SRL · CoNLL-2005 WSJ | F1: 86.0 -> 86.3 | Helped |
| Use the unified end-to-end model for span SRL | Span SRL · CoNLL-2005 Brown | F1: 76.1 -> 76.4 | Helped |
| Use the unified end-to-end model for span SRL | Span SRL · CoNLL-2012 OntoNotes | F1: 82.9 -> 83.1 | Helped |
| Use the unified end-to-end model for dependency SRL | Dependency SRL · CoNLL-2008 WSJ | F1: 85.0 -> 85.3 | Helped |
| Add ELMo representations (ablation) | Unified SRL model · CoNLL-2005 WSJ | F1: 83.0 -> 86.3 | Helped |
| Add ELMo representations (ablation) | Unified SRL model · CoNLL-2008 WSJ | F1: 85.1 -> 85.3 | No clear effect |
| Use a biaffine scorer instead of a feed-forward network scorer (ablation) | Unified SRL model · CoNLL-2005 WSJ | F1: 85.8 -> 86.3 | Helped |
| Use a biaffine scorer instead of a feed-forward network scorer (ablation) | Unified SRL model · CoNLL-2008 WSJ | F1: 83.7 -> 85.3 | Helped |
| Add self-attention to the model (ablation) | Unified SRL model · Span and dependency SRL | performance | No clear effect |
| Enforce continuation and reference role constraints during decoding (ablation) | Unified SRL model · Span and dependency SRL | performance | Hurt |
| Represent arguments as dependency heads rather than spans | Same SRL system and dependency-style evaluation · CoNLL-2005/2009 WSJ | F1: 89.20 -> 90.41 | Helped |
Closest papers by meaning
- End-to-end Semantic Role Labeling with Neural Transition-based Model (AAAI 2021 · similarity 0.91)
- Span-Based Semantic Role Labeling with Argument Pruning and Second-Order Inference (AAAI 2022 · similarity 0.91)
- Dependency Tree Representations of Predicate-Argument Structures (AAAI 2016 · similarity 0.90)
- Frame Semantic Role Labeling Using Arbitrary-Order Conditional Random Fields (AAAI 2024 · similarity 0.87)
- MAPLE: Many-Shot Adaptive Pseudo-Labeling for In-Context Learning (ICML 2025 · similarity 0.81)
- Synergistic Anchored Contrastive Pre-training for Few-Shot Relation Extraction (AAAI 2024 · 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.