Research map / Vision-language and multimodal / Cross-modal learning
Multi-Modal Fact Knowledge Generation for Imbalanced Cross-Source Entity Alignment
Qian Li, Cheng Ji, Zhaoji Liang, Yuzheng Zhang, Zhuo Chen, Siyuan Liang
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 |
|---|---|---|---|
| add LLM-based knowledge completion (ablation) | LLMEA · FB15K-DB15K (80% seeds) | MRR: 82.3 -> 85.1 | Improved |
| add LLM-based knowledge completion (ablation) | LLMEA · DBP15K ZH-EN | MRR: 87.3 -> 89.8 | Improved |
| add neighbor entity completion (ablation) | LLMEA · FB15K-DB15K (80% seeds) | MRR: 83.4 -> 85.1 | Improved |
| add relationship attribute completion (ablation) | LLMEA · FB15K-DB15K (80% seeds) | MRR: 83.0 -> 85.1 | Improved |
| add attribute similarity loss (ablation) | LLMEA · FB15K-DB15K (80% seeds) | MRR: 83.1 -> 85.1 | Improved |
| add image attributes (ablation) | LLMEA · FB15K-DB15K (80% seeds) | MRR: 82.4 -> 85.1 | Improved |
| add text attributes (ablation) | LLMEA · FB15K-DB15K (80% seeds) | MRR: 82.5 -> 85.1 | Improved |
| use LLMEA instead of DESAlign | multi-modal entity alignment · FB15K-DB15K (20% seeds) | MRR: 66.5 -> 69.2 | Improved |
| use LLMEA instead of MoAlign | multi-modal entity alignment · DBP15K ZH-EN | Hits@1: 84.7 -> 85.2 | Improved |
| completion makes alignment more robust to missing attributes, interference and attribute-count gaps | LLMEA · FB15K-DB15K and FB15K-YAGO15K | MRR | Improved |
Closest papers by meaning
- Tackling Uncertain Correspondences for Multi-Modal Entity Alignment (NeurIPS 2024 · similarity 0.89)
- HLMEA: Unsupervised Entity Alignment Based on Hybrid Language Models (AAAI 2025 · similarity 0.88)
- Tokenization, Fusion, and Augmentation: Towards Fine-grained Multi-modal Entity Representation (AAAI 2025 · similarity 0.87)
- Entity Alignment with Noisy Annotations from Large Language Models (NeurIPS 2024 · similarity 0.86)
- Pseudo-Label Calibration Semi-supervised Multi-Modal Entity Alignment (AAAI 2024 · similarity 0.86)
- MESED: A Multi-Modal Entity Set Expansion Dataset with Fine-Grained Semantic Classes and Hard Negative Entities (AAAI 2024 · similarity 0.85)
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.
