Identity, Toxicity, or Complexity? A Language-Specific Feature Selection Approach to Reclamatory Intent Detection
Published in Proceedings of EVALITA 2026 (Task A: MultiPRIDE), Bari, Italy, 2026
This paper introduces a hybrid NLP framework combining BERT embeddings with language-specific sociolinguistic features for detecting reclaimed slurs across multiple languages. The system ranked first in the Italian subtask and achieved a Macro F1 of 0.8981.
Recommended citation: Rajabi, H., Ghavidel, F., & Ghahremani, K. (2026). "Identity, Toxicity, or Complexity? A Language-Specific Feature Selection Approach to Reclamatory Intent Detection." Proceedings of EVALITA 2026 (Task A: MultiPRIDE), Bari, Italy.
Download Paper