Research map / Architectures and efficiency / Pruning and efficient tuning
Improving the Efficiency and Effectiveness for BERT-based Entity Resolution
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 delayed, enhanced alignment instead of standard pairwise BERT interaction | BERT-based entity matching · Amazon-Google | F1 (%): 73.1 -> 75.3 | Helped |
| use delayed, enhanced alignment instead of standard pairwise BERT interaction | BERT-based entity matching · iTunes-Amazon | F1 (%): 93.1 -> 96.4 | Helped |
| add comparison functions for bilateral fine-grained tuple comparison (ablation) | BERT-ER · BeerAdvo-RateBeer | F1 (%): 61.1 -> 87.5 | Helped |
| add comparison functions for bilateral fine-grained tuple comparison (ablation) | BERT-ER · Amazon-Google | F1 (%): 62.4 -> 75.3 | Helped |
| add multi-gram convolution kernels (ablation) | BERT-ER · BeerAdvo-RateBeer | F1 (%): 84.2 -> 87.5 | Helped |
| add attribute embeddings (ablation) | BERT-ER · BeerAdvo-RateBeer | F1 (%): 83.3 -> 87.5 | Helped |
| add pre-training (ablation) | BERT-ER · Amazon-Google | F1 (%): 65.8 -> 75.3 | Helped |
| jointly train blocking and entity matching | BERT-ER · Walmart-Amazon (B) | F1 (%): 94.2 -> 94.9 | Helped |
| jointly train blocking and entity matching | BERT-ER · Walmart-Amazon (B) | Pair Completeness (PC): 94.1 -> 96.1 | Helped |
| use SVD-based hyperplane orthogonalization instead of LSH | blocking · Walmart-Amazon (B) | Pair Completeness (PC): roughly 9 pts | Helped |
| integrate blocking to reduce exhaustive pair matching | BERT-ER · Walmart-Amazon (B) | decoding time: 304× less time | Helped |
| integrate blocking to reduce exhaustive pair matching | BERT-ER · Amazon-Google (B) | decoding time: 219× less time | Helped |
Closest papers by meaning
- Multipartite Entity Resolution: Motivating a K-Tuple Perspective (Student Abstract) (AAAI 2024 · similarity 0.85)
- Entity Alignment with Noisy Annotations from Large Language Models (NeurIPS 2024 · similarity 0.84)
- Knowledge Graph Completion with Relation-Aware Anchor Enhancement (AAAI 2025 · similarity 0.83)
- Select and Augment: Enhanced Dense Retrieval Knowledge Graph Augmentation (Abstract Reprint) (AAAI 2024 · similarity 0.83)
- HLMEA: Unsupervised Entity Alignment Based on Hybrid Language Models (AAAI 2025 · similarity 0.83)
- Lambda: Learning Matchable Prior For Entity Alignment with Unlabeled Dangling Cases (NeurIPS 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.