AI Paper Atlas

Research map / Large language models / Retrieval and knowledge

How to Ask Better Questions? A Large-Scale Multi-Domain Dataset for Rewriting Ill-Formed Questions

Zewei Chu, Mingda Chen, Jing Chen, Miaosen Wang, Kevin Gimpel, Manaal Faruqui, Xiance Si

AAAI 2020 · Large language models › Retrieval and knowledge · approach: Explanatory and multiple-choice QA benchmarks · 14 citations (OpenAlex)

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 changedWhereMeasuredEffect
train transformer on MQR instead of using GECquestion rewriting · MQR TESTBLEU-4: 13.1 -> 22.1Improved
use transformer instead of LSTM seq2seqMQR-trained rewriter · MQR TESTBLEU-4: 19.2 -> 22.1Improved
add Quora duplicate pairs to training (ablation)Transformer (MQR) · MQR TESTBLEU-4: 22.1 -> 23.6Improved
add Paralex pairs to training (ablation)Transformer (MQR) · MQR TESTBLEU-4: 22.1 -> 21.7No clear effect
add Quora and Paralex pairs (ablation)Transformer (MQR) · MQR TESTBLEU-4: 22.1 -> 23.1Improved
add identity well-formed pairs (ablation)Transformer (MQR) · MQR TESTBLEU-4/ROUGE-L: BLEU-4 22.1 -> 21.1; ROUGE-L 56.6 -> 58.0Mixed
run GEC after transformer (ablation)Transformer (MQR + Quora) · MQR TESTBLEU-4: 23.6 -> 26.3Improved
run GEC before transformer (ablation)Transformer (MQR + Quora) · MQR TESTBLEU-4: 23.6 -> 24.8Improved
apply transformer twice (ablation)Transformer (MQR) · MQR TESTBLEU-4: 22.1 -> 20.4Hurt
add GEC after transformer (ablation)Transformer (MQR + Quora) · MQR TEST human evalgrammar score: 0.75 -> 0.91Improved
use MQR-trained transformer instead of GECquestion rewriting · MQR TEST human evalsemantic equivalence wrt ill-formed: 1.00 -> 0.67Hurt

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy