AI RESEARCH
Multilingual Multi-Label Emotion Classification at Scale with Synthetic Data
arXiv CS.CL
•
ArXi:2604.12633v1 Announce Type: new Emotion classification in multilingual settings remains constrained by the scarcity of annotated data: existing corpora are predominantly English, single-label, and cover few languages. We address this gap by constructing a large-scale synthetic