Research map / Label-efficient and robust learning / Domain adaptation and generalization
Low Resource Sequence Tagging with Weak Labels
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 VCS to train a DNN tagger inside the ensemble (ablation) | BSC-seq · NER crowd | F1: 77.4 -> 78.0 | Helped |
| add VCS to train a DNN tagger inside the ensemble (ablation) | BSC-seq · PICO crowd | F1: 72.8 -> 77.5 | Helped |
| integrated VCS instead of aggregate-then-train DNN pipeline | BSC-seq→DNN · PICO crowd | F1: 75.5 -> 77.5 | Helped |
| integrated VCS instead of aggregate-then-train DNN pipeline | BSC-seq→DNN · NER crowd | cross entropy error: 11.0 -> 0.99 | Helped |
| add VCS to BSC-seq (ablation) | BSC-seq · NER crowd | cross entropy error: 0.65 -> 0.99 | Hurt |
| model sequential dependencies in annotator model (BSC-seq vs HMM-crowd) | HMM-crowd · PICO crowd | F1: 71.0 -> 72.8 | Helped |
| learn annotator reliability (BSC-seq vs majority vote) | MV · FAMULUS Med | mean relaxed span F1: 38.7 -> 54.2 | Helped |
| include CRFs as well as DNNs in the BSC-seq ensemble (ablation) | BSC-seq (DNN) · FAMULUS Med | mean relaxed span F1: 53.1 -> 54.2 | Helped |
| include CRFs as well as DNNs in the BSC-seq ensemble (ablation) | BSC-seq (DNN) · FAMULUS TEd | mean relaxed span F1: 43.6 -> 46.6 | Helped |
| combine out-of-domain models with BSC-seq instead of using a single DNN | Single DNN · FAMULUS TEd | mean relaxed span F1: 32.4 -> 46.6 | Helped |
| add small amounts of target-domain gold labels to BSC-seq/IBCC (ablation) | BSC-seq · FAMULUS | relaxed span F1 | Helped |
Closest papers by meaning
- Unbiased Multi-Label Learning from Crowdsourced Annotations (ICML 2024 · similarity 0.87)
- TLLC: Transfer Learning-based Label Completion for Crowdsourcing (ICML 2025 · similarity 0.85)
- Learning from Noisy Labels via Conditional Distributionally Robust Optimization (NeurIPS 2024 · similarity 0.84)
- Coupled Confusion Correction: Learning from Crowds with Sparse Annotations (AAAI 2024 · similarity 0.84)
- Multi-Instance Multi-Label Classification from Crowdsourced Labels (AAAI 2025 · similarity 0.84)
- Adaptive Integration of Partial Label Learning and Negative Learning for Enhanced Noisy Label Learning (AAAI 2024 · similarity 0.84)
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.