Research map / Label-efficient and robust learning / Recognition and long-tail
A Joint Model for Definition Extraction with Syntactic Connection and Semantic Consistency
Amir Veyseh, Franck Dernoncourt, Dejing Dou, Thien Nguyen
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 changed | Where | Measured | Effect |
|---|---|---|---|
| jointly add sentence classification to sequence labeling (ablation) | Ours · Textbook and Contract | F1: 48.8 -> 50.6; 68.5 -> 71.7 | Improved |
| add GCN over the dependency tree (ablation) | Ours · Textbook and Contract | F1: 48.8 -> 50.6; 68.5 -> 71.7 | Improved |
| add dependency path prediction (ablation) | Ours · Textbook and Contract | F1: 48.1 -> 50.6; 68.2 -> 71.7 | Improved |
| add direct local semantic consistency (ablation) | Ours · Textbook and Contract | F1: 49.6 -> 50.6; 70.5 -> 71.7 | Improved |
| add indirect local semantic consistency (ablation) | Ours · Textbook and Contract | F1: 49.8 -> 50.6; 69.6 -> 71.7 | Improved |
| add global semantic consistency (ablation) | Ours · Textbook and Contract | F1: 49.0 -> 50.6; 69.1 -> 71.7 | Improved |
| remove DPP and increase GCN from 2 to 3 layers (ablation) | Ours · Textbook and Contract | F1: 50.6 -> 48.5; 71.7 -> 68.0 | Hurt |
| remove DPP and increase GCN from 2 to 4 layers (ablation) | Ours · Textbook and Contract | F1: 50.6 -> 48.0; 71.7 -> 67.3 | Hurt |
| use the joint model instead of CVT | Sequence labeling · WCL, W00, Textbook, and Contract | F1: 82.5 -> 83.3; 58.9 -> 60.6; 47.3 -> 50.6; 68.2 -> 71.7 | Improved |
| use the joint model instead of SA | Sequence classification · WCL, W00, Textbook, and Contract | F1: 94.2 -> 99.5; 57.4 -> 67.2; 64.2 -> 70.3; 84.1 -> 91.7 | Improved |
| initialize embeddings with BERT | Ours · WCL and W00 sequence labeling | F1: 83.3 -> 85.3; 60.6 -> 66.9 | Improved |
| initialize embeddings with BERT | Ours · Textbook sequence labeling | F1: 50.6 -> 54.0 | Improved |
Closest papers by meaning
- Definition Modeling: Learning to Define Word Embeddings in Natural Language (AAAI 2017 · similarity 0.87)
- Multi-View Consistency for Relation Extraction via Mutual Information and Structure Prediction (AAAI 2020 · similarity 0.85)
- SEE: Syntax-Aware Entity Embedding for Neural Relation Extraction (AAAI 2018 · similarity 0.84)
- Learning to Map Frequent Phrases to Sub-Structures of Meaning Representation for Neural Semantic Parsing (AAAI 2020 · similarity 0.84)
- Neural Relation Extraction within and across Sentence Boundaries (AAAI 2019 · similarity 0.83)
- Narrowing the Gap between Supervised and Unsupervised Sentence Representation Learning with Large Language Model (AAAI 2024 · similarity 0.82)
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.
