Research map / Label-efficient and robust learning / Domain adaptation and generalization
Unsupervised Cross-Domain Rumor Detection with Contrastive Learning and Cross-Attention
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 UCD-RD instead of UCD-CEloss | Cross-domain rumor detection · Terrorist→Gossip | accuracy: 74.01 -> 84.88 | Helped |
| Use UCD-RD instead of UCD-CEloss | Cross-domain rumor detection · Twitter→Twitter_COVID19 | accuracy: 58.87 -> 66.50 | Helped |
| Use UCD-RD instead of UCD-CEloss | Cross-domain rumor detection · Twitter15→Twitter16 | accuracy: 75.15 -> 79.47 | Helped |
| Use UCD-RD instead of UCD-CEloss | Cross-domain rumor detection · Weibo→Weibo_COVID19 | accuracy: 57.96 -> 68.92 | Helped |
| Add source in-domain instance contrastive loss (ablation) | UCD-RD · Four cross-domain dataset groups | accuracy: 74.01 -> 74.58; 58.87 -> 58.95; 75.15 -> 75.17; 57.96 -> 58.16 | Helped |
| Add target in-domain instance contrastive loss (ablation) | UCD-RD · Four cross-domain dataset groups | accuracy: 74.58 -> 75.65; 58.95 -> 59.78; 75.17 -> 76.32; 58.16 -> 60.03 | Helped |
| Add target-to-source instance cross-domain contrastive loss (ablation) | UCD-RD · Four cross-domain dataset groups | accuracy: 75.65 -> 77.90; 59.78 -> 61.55; 76.32 -> 77.23; 60.03 -> 63.16 | Helped |
| Add source-to-target instance cross-domain contrastive loss (ablation) | UCD-RD · Four cross-domain dataset groups | accuracy: 77.90 -> 79.01; 61.55 -> 62.25; 77.23 -> 77.89; 63.16 -> 65.23 | Helped |
| Add target-to-source prototype-wise contrastive loss (ablation) | UCD-RD · Four cross-domain dataset groups | accuracy: 79.01 -> 79.58; 62.25 -> 63.77; 77.89 -> 78.23; 65.23 -> 65.79 | Helped |
| Add cross-attention loss (ablation) | UCD-RD · Four cross-domain dataset groups | accuracy: 79.58 -> 84.88; 63.77 -> 66.50; 78.23 -> 79.47; 65.79 -> 68.92 | Helped |
Closest papers by meaning
- Propagation Tree Is Not Deep: Adaptive Graph Contrastive Learning Approach for Rumor Detection (AAAI 2024 · similarity 0.86)
- Frequency Spectrum Is More Effective for Multimodal Representation and Fusion: A Multimodal Spectrum Rumor Detector (AAAI 2024 · similarity 0.84)
- DAMMFND: Domain-Aware Multimodal Multi-view Fake News Detection (AAAI 2025 · similarity 0.83)
- Text-Guided Fine-grained Counterfactual Inference for Short Video Fake News Detection (AAAI 2025 · similarity 0.83)
- Bad Actor, Good Advisor: Exploring the Role of Large Language Models in Fake News Detection (AAAI 2024 · similarity 0.82)
- External Reliable Information-enhanced Multimodal Contrastive Learning for Fake News Detection (AAAI 2025 · similarity 0.81)
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.