Research map / Label-efficient and robust learning / Domain adaptation and generalization
Transfer Learning for Cross-Language Text Categorization through Active Correspondences Construction
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 |
|---|---|---|---|
| add MMD distribution-matching constraint to matrix completion (DMMC vs TSL) (ablation) | matrix-completion HTL · cross-language sentiment (9 tasks) | accuracy: B-FR 73.95 -> 76.52 | Helped |
| add MMD distribution-matching constraint (DMMC vs TSL) (ablation) | matrix-completion HTL · Reuters topic categorization | accuracy: SP 56.98 -> 62.64 | Helped |
| use DMMC instead of TSL with few correspondences (ablation) | matrix-completion HTL · sentiment and topic, varying correspondence size | accuracy | Helped |
| choose correspondences actively instead of randomly (ablation) | DMMC · sentiment and Reuters, k=100 | accuracy | Helped |
| fill missing entries by matrix completion before dimensionality reduction (ablation) | LSI · sentiment and topic | accuracy: LSI B-FR 71.56 vs DMMC 76.52 | Helped |
| decrease low-rank weight gamma to 1 or below (ablation) | DMMC · sentiment (FR, GE, JP) | accuracy | Hurt |
| increase sparsity weight mu to large values (ablation) | DMMC · sentiment (FR, GE, JP) | accuracy | Hurt |
| compare SVM-SC with OPCA, LSI and KCCA | SVM-SC · Reuters GE | accuracy: FR 63.77 vs 58.70; GE 47.59 vs 53.18 (OPCA) | Mixed |
Closest papers by meaning
- Exploiting Cross-Lingual Subword Similarities in Low-Resource Document Classification (AAAI 2020 · similarity 0.86)
- Active Learning with Cross-Class Similarity Transfer (AAAI 2017 · similarity 0.85)
- Harnessing Language Model for Cross-Heterogeneity Graph Knowledge Transfer (AAAI 2025 · similarity 0.84)
- MMTEB: Massive Multilingual Text Embedding Benchmark (ICLR 2025 · similarity 0.83)
- Constrained Decoding for Cross-lingual Label Projection (ICLR 2024 · similarity 0.82)
- RMIB: Representation Matching Information Bottleneck for Matching Text Representations (ICML 2024 · 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.