Research map / Graph learning / Knowledge and temporal graphs
A Neural Transition-Based Approach for Semantic Dependency Graph Parsing
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 Bi-LSTM Subtraction buffer representation (ablation) | basic Dyer-style parser · CSDG NEWS | LF: 62.23 -> 62.59 | Helped |
| add Incremental Tree-LSTM for sub-graphs (ablation) | basic Dyer-style parser · CSDG NEWS | LF: 62.23 -> 62.62 | Helped |
| add both Bi-LSTM Subtraction and Incremental Tree-LSTM (ablation) | basic Dyer-style parser · CSDG NEWS | LF: 62.23 -> 63.30 | Helped |
| add both modules (ablation) | basic Dyer-style parser · CSDG NEWS | NLF: 49.18 -> 51.16 | Helped |
| add both modules (ablation) | basic Dyer-style parser · CSDG TEXTBOOKS | LF: 71.51 -> 72.92 | Helped |
| add both modules (ablation) | basic Dyer-style parser · CSDG TEXTBOOKS | NLF: 59.70 -> 61.91 | Helped |
| add Bi-LSTM Subtraction and Incremental Tree-LSTM (ablation) | basic parser · CSDG TEXTBOOK | speed tokens/s: 335.0 -> 258.1 | Hurt |
| update all ancestors in Tree-LSTM instead of only the head (ablation) | BS-IT · CSDG | LF/NLF | No clear effect |
| update all ancestors in Tree-LSTM (ablation) | BS-IT · CSDG TEXTBOOK | speed tokens/s: 258.1 -> 229.7 | Hurt |
| ignore dependency relations in RecNN (ablation) | basic system · CSDG | LF | No clear effect |
| one-stage graph parser vs two-stage tree-then-classifier | BS-IT vs Ding et al. 2-stage · CSDG NEWS | NLF: 39.93 -> 51.16 | Helped |
| ensemble of 3 models by summing transition scores (ablation) | BS-IT · SemEval-2015 English in-domain | Avg labeled F1: 85.6 -> 86.9 | Helped |
Closest papers by meaning
- Discontinuous Constituent Parsing with Pointer Networks (AAAI 2020 · similarity 0.89)
- Global Greedy Dependency Parsing (AAAI 2020 · similarity 0.89)
- Dependency Parsing is More Parameter-Efficient with Normalization (NeurIPS 2025 · similarity 0.87)
- Dependency Structure-Enhanced Graph Attention Networks for Event Detection (AAAI 2024 · similarity 0.85)
- Implicit Word Reordering with Knowledge Distillation for Cross-Lingual Dependency Parsing (AAAI 2025 · similarity 0.84)
- Tracing Hyperparameter Dependencies for Model Parsing via Learnable Graph Pooling Network (NeurIPS 2024 · similarity 0.84)
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.