Research map / Detection and segmentation / CNN features
A Unified Multi-Task Learning Framework for Joint Extraction of Entities and Relations
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 the full UMT framework with pseudo-question-generated object queries | joint entity and relation extraction · CoNLL04 | Relation F1: 71.5 -> 72.4 | Helped |
| add global relation prediction to local relation prediction (ablation) | UMT w/ PseudoGQ · CoNLL04 | Relation F1: 72.2 -> 72.4 | Helped |
| add global and local subject-aware relation prediction (ablation) | UMT w/ PseudoGQ · CoNLL04 | Relation F1: 71.8 -> 72.4 | Helped |
| add the context fusion layer (ablation) | UMT w/ PseudoGQ · CoNLL04 | Relation F1: 71.3 -> 72.4 | Helped |
| use QA-based object extraction instead of multi-head selection (ablation) | UMT with SRP · CoNLL04 | Relation F1: 64.7 -> 72.4 | Helped |
| add global relation prediction to local relation prediction (ablation) | UMT w/ PseudoGQ · ACE05 | Relation F1: 61.1 -> 61.2 | Helped |
| add global and local subject-aware relation prediction (ablation) | UMT w/ PseudoGQ · ACE05 | Relation F1: 60.3 -> 61.2 | Helped |
| add the context fusion layer (ablation) | UMT w/ PseudoGQ · ACE05 | Relation F1: 60.7 -> 61.2 | Helped |
| use QA-based object extraction instead of multi-head selection (ablation) | UMT with SRP · ACE05 | Relation F1: 59.4 -> 61.2 | Helped |
| use relation-middle extraction rather than relation-first extraction | UMT · DuIE2.0 | Relation F1: 75.5 -> 64.9 | Hurt |
| use relation-middle extraction rather than relation-first extraction | UMT · CoNLL04 | Relation F1: 68.0 -> 72.4 | Helped |
| use NL-question-driven rather than pseudo-question-driven generated queries | UMT · Travel20 non-predefined relation extraction | F1-score: 80.1 -> 88.2 | Helped |
Closest papers by meaning
- Joint Extraction of Entities and Overlapping Relations Using Position-Attentive Sequence Labeling (AAAI 2019 · similarity 0.90)
- Entity Structure Within and Throughout: Modeling Mention Dependencies for Document-Level Relation Extraction (AAAI 2021 · similarity 0.90)
- Distant Supervision for Relation Extraction with Sentence-Level Attention and Entity Descriptions (AAAI 2017 · similarity 0.89)
- A Hierarchical Network for Multimodal Document-Level Relation Extraction (AAAI 2024 · similarity 0.87)
- DocKS-RAG: Optimizing Document-Level Relation Extraction through LLM-Enhanced Hybrid Prompt Tuning (ICML 2025 · similarity 0.87)
- Prototype-Guided Multimodal Relation Extraction based on Entity Attributes (AAAI 2025 · 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.