Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Neural Semantic Parsing in Low-Resource Settings with Back-Translation and Meta-Learning
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 back-translation, quality control, and MAML to the rule-initialized parser | WikiSQL parser · WikiSQL test | execution accuracy: 70.3% -> 72.7% | Helped |
| Add question generation to rule-covered training (ablation) | WikiSQL parser · WikiSQL test | execution accuracy: 70.3% -> 72.5% | Helped |
| Add back-translation to rule-covered training (ablation) | WikiSQL parser · WikiSQL test | execution accuracy: 70.3% -> 72.3% | Helped |
| Add quality control to back-translation (ablation) | WikiSQL parser · WikiSQL test | execution accuracy: 72.3% -> 72.6% | Helped |
| Add MAML to back-translation with quality control (ablation) | WikiSQL parser · WikiSQL test | execution accuracy: 72.6% -> 72.7% | Helped |
| Use question generation instead of self-training (ablation) | WikiSQL parser · WikiSQL test | execution accuracy: 70.3% -> 72.5% | Helped |
| Add back-translation, quality control, and MAML to the rule-initialized parser | SimpleQuestions parser · SimpleQuestions test | accuracy: 48.0% -> 58.1% | Helped |
| Add back-translation to rule-covered training (ablation) | SimpleQuestions parser · SimpleQuestions test | accuracy: 48.0% -> 57.1% | Helped |
| Add quality control to back-translation (ablation) | SimpleQuestions parser · SimpleQuestions test | accuracy: 57.1% -> 57.6% | Helped |
| Add MAML to back-translation with quality control (ablation) | SimpleQuestions parser · SimpleQuestions test | accuracy: 57.6% -> 58.1% | Helped |
| Use question generation instead of self-training (ablation) | SimpleQuestions parser · SimpleQuestions test | accuracy: 55.9% -> 52.8% | Hurt |
| Add back-translation, quality control, and MAML to the rule-initialized parser | SequentialQA parser · SequentialQA test | question answering accuracy: 34.3% -> 32.7% | Hurt |
Closest papers by meaning
- Merging Weak and Active Supervision for Semantic Parsing (AAAI 2020 · similarity 0.89)
- Learning to Map Frequent Phrases to Sub-Structures of Meaning Representation for Neural Semantic Parsing (AAAI 2020 · similarity 0.88)
- Semantic Parsing with Neural Hybrid Trees (AAAI 2017 · similarity 0.88)
- BenchCLAMP: A Benchmark for Evaluating Language Models on Syntactic and Semantic Parsing (NeurIPS 2023 · similarity 0.88)
- GraPPa: Grammar-Augmented Pre-Training for Table Semantic Parsing (ICLR 2021 · similarity 0.87)
- Filling Memory Gaps: Enhancing Continual Semantic Parsing via SQL Syntax Variance-Guided LLMs Without Real Data Replay (AAAI 2025 · similarity 0.85)
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.