Zero-shot classifiers built on sentence-transformer language models (T5-xl, MiniLM) outperform much larger LLMs on multi-label, hierarchical requirements classification, and a hop-count label distance metric helps rank models.
Zero-shot learning for require- ments classification: An exploratory study,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.SE 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Language Models to Support Multi-Label Classification of Industrial Data
Zero-shot classifiers built on sentence-transformer language models (T5-xl, MiniLM) outperform much larger LLMs on multi-label, hierarchical requirements classification, and a hop-count label distance metric helps rank models.