Research map / Vision-language and multimodal / Visual question answering
A Question-Focused Multi-Factor Attention Network for Question Answering
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 |
|---|---|---|---|
| Use AMANDA instead of R2-BiLSTM | NewsQA answer extraction · NewsQA test | F1: 56.7 -> 63.7 | Helped |
| Use AMANDA instead of ASR | SearchQA answer extraction · SearchQA test | N-gram F1: 22.8 -> 56.6 | Helped |
| Add multi-factor attentive encoding (ablation) | AMANDA · NewsQA dev | F1: 61.2 -> 63.3 | Helped |
| Add max-attentional question aggregation and question type representation (ablation) | AMANDA · NewsQA dev | F1: 60.5 -> 63.3 | Helped |
| Add max-attentional question aggregation (ablation) | AMANDA · NewsQA dev | F1: 61.3 -> 63.3 | Helped |
| Add question type representation (ablation) | AMANDA · NewsQA dev | F1: 61.8 -> 63.3 | Helped |
| Add character embeddings (ablation) | AMANDA · NewsQA dev | F1: 61.4 -> 63.3 | Helped |
| Add question-dependent passage encoding (ablation) | AMANDA · NewsQA dev | F1: 45.0 -> 63.3 | Helped |
| Use a second prediction LSTM instead of a feed-forward layer (ablation) | AMANDA · NewsQA dev | F1: 61.6 -> 63.3 | Helped |
| Use 4 attention factors rather than 1, 2, or 3 (ablation) | AMANDA · NewsQA dev | F1: 61.2; 61.9; 62.9; 63.3 | Helped |
| Increase attention factors from 4 to 5 (ablation) | AMANDA · NewsQA dev | F1: 63.3 -> 62.5 | Hurt |
| Use max rather than mean question aggregation (ablation) | AMANDA · NewsQA dev | F1: 61.3 -> 63.3 | Helped |
Closest papers by meaning
- Hybrid Attentive Answer Selection in CQA With Deep Users Modelling (AAAI 2018 · similarity 0.88)
- Stacked Attention Networks for Image Question Answering (CVPR 2016 · similarity 0.88)
- Multi-Level Attention Networks for Visual Question Answering (CVPR 2017 · similarity 0.88)
- Question-Answering with Grammatically-Interpretable Representations (AAAI 2018 · similarity 0.88)
- DSAS: A Universal Plug-and-Play Framework for Attention Optimization in Multi-Document Question Answering (NeurIPS 2025 · similarity 0.87)
- Question Calibration and Multi-Hop Modeling for Temporal 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.