Research map / Large language models / Retrieval and knowledge
Learning From Unannotated QA Pairs to Analogically Disambiguate and Answer Questions
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 |
|---|---|---|---|
| correct the 22 Geoquery gold queries that evaluate incorrectly | Our System, 10-fold · Geoquery | answer accuracy (%): 86.3 -> 88.8 | Helped |
| correct the 22 Geoquery gold queries that evaluate incorrectly | Our System, 2-fold · Geoquery | answer accuracy (%): 84.8 -> 87.4 | Helped |
| reduce cross-validation from 10-fold to 2-fold (far fewer training examples) | Our System · Geoquery corrected | answer accuracy (%): 88.8 -> 87.4 | Hurt |
| reduce training set to 10 QA pairs | Our System · Geoquery corrected | answer accuracy (%): 88.8 -> 51.6 | Hurt |
| use only 100 QA pairs for training instead of full 10-fold training | Our System · Geoquery corrected | answer accuracy (%): 88.8 -> 85.6 | No clear effect |
| compare unannotated-QA training against annotated logical-form training (KZGS10) | Our System corrected 10-fold · Geoquery | answer accuracy (%): 88.9 -> 88.8 | No clear effect |
| LJK11 manually augmented triggers versus base triggers | LJK11, 680/200 · Geoquery | answer accuracy (%): 87.9 -> 91.4 | Helped |
Closest papers by meaning
- Question Answering as Global Reasoning Over Semantic Abstractions (AAAI 2018 · similarity 0.89)
- Open-Vocabulary Semantic Parsing with both Distributional Statistics and Formal Knowledge (AAAI 2017 · similarity 0.88)
- Complex Sequential Question Answering: Towards Learning to Converse Over Linked Question Answer Pairs with a Knowledge Graph (AAAI 2018 · similarity 0.87)
- SuRe: Summarizing Retrievals using Answer Candidates for Open-domain QA of LLMs (ICLR 2024 · similarity 0.86)
- Making Retrieval-Augmented Language Models Robust to Irrelevant Context (ICLR 2024 · similarity 0.86)
- FlexKBQA: A Flexible LLM-Powered Framework for Few-Shot Knowledge Base Question Answering (AAAI 2024 · similarity 0.86)
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.