AI Paper Atlas

Research map / Label-efficient and robust learning / Semi-supervised learning

Weakly-Supervised Hierarchical Text Classification

Yu Meng, Jiaming Shen, Chao Zhang, Jiawei Han

AAAI 2019 · Label-efficient and robust learning › Semi-supervised learning · approach: Self-training theory

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 changedWhereMeasuredEffect
Add WeSHClass's weak-supervision and hierarchical training approachWeSHClass · NYT, arXiv, and Yelp Review; keyword and document supervisionMacro-F1 and Micro-F1Improved
Add the global classifier ensemble instead of greedy local predictions (ablation)WeSHClass · NYT keywordsMacro-F1: 0.618 -> 0.632Improved
Add movMF class-semantic modeling for pseudo-document generation (ablation)WeSHClass · NYT keywordsMacro-F1: 0.628 -> 0.632Improved
Add global self-training (ablation)WeSHClass · NYT keywordsMacro-F1: 0.550 -> 0.632Improved
Add the global classifier ensemble instead of greedy local predictions (ablation)WeSHClass · NYT docsMacro-F1: 0.520 (0.065) -> 0.532 (0.015)Improved
Add movMF class-semantic modeling for pseudo-document generation (ablation)WeSHClass · NYT docsMacro-F1: 0.527 (0.031) -> 0.532 (0.015)Improved
Add global self-training (ablation)WeSHClass · NYT docsMacro-F1: 0.491 (0.036) -> 0.532 (0.015)Improved
Generate pseudo documents with movMF plus an LSTM language model rather than bag-of-words or bag-of-words plus reorderingWeSHClass · NYT keywords, level 1 self-trainingperformanceImproved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy