Research map / Label-efficient and robust learning / Recognition and long-tail
Hierarchical Text Classification as Sub-hierarchy Sequence Generation
SangHun Im, GiBaeg Kim, Heung-Seon Oh, Seongung Jo, Dong Hwan Kim
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 |
|---|---|---|---|
| Use sub-hierarchy sequence generation instead of encoding the entire hierarchy in the model structure | HiDEC · RCV1-v2 | Micro-F1: 84.73 -> 85.54 | Improved |
| Use sub-hierarchy sequence generation instead of encoding the entire hierarchy in the model structure | HiDEC · NYT | Macro-F1: 60.83 -> 63.99 | Improved |
| Use sub-hierarchy sequence generation instead of encoding the entire hierarchy in the model structure | HiDEC · EURLEX57K | Micro-F1: 71.11 -> 71.23 | Improved |
| Add a residual connection (ablation) | HiDEC · RCV1-v2 | Micro-F1: 85.54 -> 84.69 | Hurt |
| Add a residual connection (ablation) | HiDEC · EURLEX57K | Micro-F1: 71.31 -> 69.97 | Hurt |
| Remove hierarchy-aware masking (ablation) | HiDEC · RCV1-v2 | Micro-F1: 85.54 -> 85.46 | Hurt |
| Remove hierarchy-aware masking (ablation) | HiDEC · EURLEX57K | Micro-F1: 71.31 -> 71.22 | Hurt |
| Remove level embedding (ablation) | HiDEC · RCV1-v2 | Micro-F1: 85.54 -> 85.51 | Hurt |
| Remove level embedding (ablation) | HiDEC · EURLEX57K | Micro-F1: 71.31 -> 70.90 | Hurt |
Closest papers by meaning
- Weakly-Supervised Hierarchical Text Classification (AAAI 2019 · similarity 0.88)
- Hierarchical Multi-Label Classification Networks (ICML 2018 · similarity 0.85)
- Self-Paced Unified Representation Learning for Hierarchical Multi-Label Classification (AAAI 2024 · similarity 0.83)
- LayoutFormer: Hierarchical Text Detection Towards Scene Text Understanding (CVPR 2024 · similarity 0.81)
- On the Affinity, Rationality, and Diversity of Hierarchical Topic Modeling (AAAI 2024 · similarity 0.81)
- Language Models as Hierarchy Encoders (NeurIPS 2024 · 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.
