Research map / Vision-language and multimodal / CLIP and zero-shot
A Bidirectional Multi-paragraph Reading Model for Zero-shot Entity Linking
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 Whole Entity Masking pre-training (ablation) | BERT baseline · Validation set | Macro accuracy: 78.07 -> 78.18 | Helped |
| Add Whole Entity Masking pre-training (ablation) | BERT baseline · Test set | Macro accuracy: 76.38 -> 77.14 | Helped |
| Add Whole Entity Masking pre-training (ablation) | BERT baseline · Low Overlap test subset | accuracy: 71.46 -> 71.03 | Hurt |
| Add multi-paragraph reading of entity descriptions (ablation) | BERT baseline · Validation set | Macro accuracy: 78.07 -> 79.48 | Helped |
| Add multi-paragraph reading of entity descriptions (ablation) | BERT baseline · Test set | Macro accuracy: 76.38 -> 77.62 | Helped |
| Add backward matching over mention-document paragraphs (ablation) | Uni-MPR · Validation set | Macro accuracy: 80.81 -> 81.98 | Helped |
| Add backward matching over mention-document paragraphs (ablation) | Uni-MPR · Test set | Macro accuracy: 78.42 -> 79.97 | Helped |
| Increase the number of paragraphs while holding paragraph lengths fixed (ablation) | Uni-MPR and Bi-MPR · Validation set | accuracy | Helped |
| Use shorter paragraphs and more paragraphs instead of longer paragraphs at the same total input length (ablation) | Bi-MPR · Validation set | accuracy | Hurt |
| Use longer paragraphs in the initial mention-entity cross-attention (ablation) | Bi-MPR · Validation set | inference time | Hurt |
Closest papers by meaning
- A Dual-Way Enhanced Framework from Text Matching Point of View for Multimodal Entity Linking (AAAI 2024 · similarity 0.86)
- Multi-Grained Query-Guided Set Prediction Network for Grounded Multimodal Named Entity Recognition (AAAI 2025 · similarity 0.84)
- Zero-Shot Image Captioning with Multi-type Entity Representations (AAAI 2025 · similarity 0.84)
- Span Graph Transformer for Document-Level Named Entity Recognition (AAAI 2024 · similarity 0.84)
- A Hierarchical Network for Multimodal Document-Level Relation Extraction (AAAI 2024 · similarity 0.84)
- Revisiting Document-Level Relation Extraction with Context-Guided Link Prediction (AAAI 2024 · 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.