Research map / Label-efficient and robust learning / Recognition and long-tail
Constructing Superior Representations Beyond the Original Documents via a Contrastive Gaussian Fusion Network for Clustering
Ao Shen, Ruizhang Huang, Jingjing Xue, Ruina Bai
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 consistency enhancement (beta<1 KL weight) in place of standard VAE regularization (ablation) | CGFN without clustering layer · BBC | NMI: 51.20 -> 63.45 | Improved |
| add consistency enhancement in place of standard VAE regularization (ablation) | CGFN without clustering layer · Reuters-10k | NMI: 43.28 -> 54.09 | Improved |
| add contrastive Gaussian fusion with neighbor data to single-input VAE (ablation) | CGFN-f · Abstract | NMI: 33.02 -> 45.75 | Improved |
| add contrastive Gaussian fusion with neighbor data to single-input VAE (ablation) | CGFN-f · BBCSports | NMI: 34.05 -> 50.91 | Improved |
| add clustering layer (self-training KL objective) (ablation) | CGFN-c · BBCSports | NMI: 58.40 -> 85.79 | Improved |
| use CGFN instead of best baseline | document clustering · Reuters-10k | ACC: 77.15 -> 75.71 | Hurt |
| use CGFN instead of best baseline | document clustering · four datasets | NMI: +11.73, +2.81, +17.50, +3.50 points on BBC, Abstract, BBCSports, Reuters-10k | Improved |
Closest papers by meaning
- THESAURUS: Contrastive Graph Clustering by Swapping Fused Gromov-Wasserstein Couplings (AAAI 2025 · similarity 0.85)
- GeCC: Generalized Contrastive Clustering with Domain Shifts Modeling (AAAI 2025 · similarity 0.84)
- Deep Contrastive Graph Learning with Clustering-Oriented Guidance (AAAI 2024 · similarity 0.83)
- Improved Graph Contrastive Learning for Short Text Classification (AAAI 2024 · similarity 0.83)
- Contrastive Learning with Simplicial Convolutional Networks for Short-Text Classification (ICML 2025 · similarity 0.83)
- Ambiguous Instance-Aware Contrastive Network with Multi-Level Matching for Multi-View Document Clustering (AAAI 2025 · 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.
