Research map / Graph learning / Knowledge and temporal graphs
Graph-Based Transformer with Cross-Candidate Verification for Semantic Parsing
Bo Shao, Yeyun Gong, Weizhen Qi, Guihong Cao, Jianshu Ji, Xiaola Lin
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 cross-candidate verification (ablation) | GTCV · TOP test | exact match accuracy: 81.16 -> 82.51 | Improved |
| add graph-based matching (ablation) | GTCV · TOP test | exact match accuracy: 81.26 -> 82.51 | Improved |
| add graph matching and verification on top of BERT reranker | BERT reranker · TOP test | exact match accuracy: 80.15 -> 82.51 | Improved |
| add GTCV reranking to seq2seq generation | seq2seq generation · TOP test | exact match accuracy: 77.40 -> 82.51 | Improved |
| add GTCV reranking | seq2seq generation · ATIS | accuracy: 82.4 -> 87.3 | Improved |
| add GTCV reranking | seq2seq generation · JOBS | accuracy: 89.3 -> 92.9 | Improved |
| add GTCV reranking | seq2seq generation · TOP in-domain subset | exact match accuracy: 79.82 -> 85.84 | Improved |
| replace BERT reranker with GTCV | reranker · JOBS | accuracy: 90.1 -> 92.9 | Improved |
| increase candidate number from 3 to 10 for cross-candidate verification (ablation) | GTCV · TOP test | exact match accuracy: 80.01 -> 82.51 | Improved |
Closest papers by meaning
- One SPRING to Rule Them Both: Symmetric AMR Semantic Parsing and Generation without a Complex Pipeline (AAAI 2021 · similarity 0.87)
- Graph Transformer for Graph-to-Sequence Learning (AAAI 2020 · similarity 0.87)
- A Neural Transition-Based Approach for Semantic Dependency Graph Parsing (AAAI 2018 · similarity 0.87)
- Talk like a Graph: Encoding Graphs for Large Language Models (ICLR 2024 · similarity 0.85)
- KnowFormer: Revisiting Transformers for Knowledge Graph Reasoning (ICML 2024 · similarity 0.84)
- From Sequence to Structure: Uncovering Substructure Reasoning in Transformers (NeurIPS 2025 · 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.
