Research map / Label-efficient and robust learning / Semi-supervised learning
Contrast-Enhanced Semi-supervised Text Classification with Few Labels
Austin Cheng-Yun Tsai, Sheng-Ya Lin, Li-Chen Fu
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 certainty-driven sample selection (ablation) | CEST · five benchmark datasets | average test accuracy (%): 87.89 -> 91.57 | Improved |
| Add the noise-robust PHCE loss (ablation) | CEST · five benchmark datasets | average test accuracy (%): 91.04 -> 91.57 | Improved |
| Add reliable sub-graph construction (ablation) | CEST · five benchmark datasets | average test accuracy (%): 90.46 -> 91.57 | Improved |
| Add smoothness regularization (ablation) | CEST · five benchmark datasets | average test accuracy (%): 89.97 -> 91.57 | Improved |
| Use CEST instead of UST | BERT semi-supervised text classification · five benchmark datasets | average test accuracy (%): 89.46 (+5.85%) -> 91.57 (+7.96%) | Improved |
| Use CEST instead of UDA | BERT semi-supervised text classification · five benchmark datasets | average test accuracy (%): 89.06 (+5.45%) -> 91.57 (+7.96%) | Improved |
| Use CEST instead of BERT fine-tuning without unlabeled data | BERT · five benchmark datasets | average test accuracy (%): 83.61 (+0.00%) -> 91.57 (+7.96%) | Improved |
Closest papers by meaning
- Improved Graph Contrastive Learning for Short Text Classification (AAAI 2024 · similarity 0.88)
- A Simple Graph Contrastive Learning Framework for Short Text Classification (AAAI 2025 · similarity 0.86)
- Contrastive Credibility Propagation for Reliable Semi-supervised Learning (AAAI 2024 · similarity 0.86)
- Weakly-Supervised Contrastive Learning for Imprecise Class Labels (ICML 2025 · similarity 0.86)
- Boosting Short Text Classification with Multi-Source Information Exploration and Dual-Level Contrastive Learning (AAAI 2025 · similarity 0.85)
- Continuous Contrastive Learning for Long-Tailed Semi-Supervised Recognition (NeurIPS 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.
