Research map / Vision-language and multimodal / Visual question answering
Improved Text Matching by Enhancing Mutual Information
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 reformulated question (GAN rewriting) with ranking loss (ablation) | QA matching model (GAN question rewriting) · SemEval 2016 Task 3A | MAP: 78.03 -> 80.36 | Helped |
| add reformulated question with ranking loss (ablation) | QA matching model (GAN question rewriting) · WikiQA | MAP: 72.16 -> 73.48 | Helped |
| add reformulated question with ranking loss (ablation) | QA matching model (GAN question rewriting) · ZhihuRank | NDCG: 0.7357 -> 0.7723 | Helped |
| add reformulated question with ranking loss (ablation) | QA matching model (GAN question rewriting) · ZhihuRank | ERR: 0.6345 -> 0.6502 | Helped |
| add reformulated question with classification loss (ablation) | QA matching model (GAN question rewriting) · SemEval 2016 Task 3A | MAP: 75.86 -> 77.62 | Helped |
| add reformulated question with classification loss (ablation) | QA matching model (GAN question rewriting) · WikiQA | MAP: 69.92 -> 71.83 | Helped |
| add reformulated question with classification loss (ablation) | QA matching model (GAN question rewriting) · ZhihuRank | NDCG: 0.7028 -> 0.7451 | Helped |
| use LambdaRank listwise loss instead of classification loss (with rewriting) (ablation) | QA matching model (GAN question rewriting) · SemEval 2016 Task 3A | MAP: 77.62 -> 80.36 | Helped |
| use LambdaRank listwise loss instead of classification loss (with rewriting) (ablation) | QA matching model (GAN question rewriting) · WikiQA | MAP: 71.83 -> 73.48 | Helped |
| use LambdaRank listwise loss instead of classification loss (with rewriting) (ablation) | QA matching model (GAN question rewriting) · ZhihuRank | NDCG: 0.7451 -> 0.7723 | Helped |
Closest papers by meaning
- Knowledge and Cross-Pair Pattern Guided Semantic Matching for Question Answering (AAAI 2020 · similarity 0.87)
- Knowledge Enhanced Hybrid Neural Network for Text Matching (AAAI 2018 · similarity 0.86)
- UniGen: A Unified Generative Framework for Retrieval and Question Answering with Large Language Models (AAAI 2024 · similarity 0.84)
- RMIB: Representation Matching Information Bottleneck for Matching Text Representations (ICML 2024 · similarity 0.84)
- Retrieval is Accurate Generation (ICLR 2024 · similarity 0.83)
- Cooperative Retrieval-Augmented Generation for Question Answering: Mutual Information Exchange and Ranking by Contrasting Layers (NeurIPS 2025 · similarity 0.83)
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.