New Label Semantic Expansion method enhances topic modeling with guided neural networks

Researchers have introduced Label Semantic Expansion (LSE), a novel approach to topic modeling that focuses on enriching sparse label representations with descriptive topic words derived from a corpus. This method is instantiated through a Label-Guided Neural Topic Model (LGNTM), which learns topics specifically aligned with labels. LGNTM grounds these topics in both lexical and document semantic spaces, ensuring consistency between topic and label structures. Experiments indicate strong performance in label-topic alignment, label expansion, topic quality, and downstream classification tasks. AI

IMPACT This research could improve the accuracy and interpretability of topic modeling for content analysis and downstream classification tasks.

RANK_REASON The item is an academic paper describing a new method for topic modeling. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• alphaXiv
• arXiv
• CatalyzeX
• Connected Papers
• CORE Recommender
• DagsHub
• Gotit.pub
• Hugging Face
• Label Guided Neural Topic Model
• Label Semantic Expansion
• LGNTM
• Litmaps
• ScienceCast
• scite Smart Citations

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →