Open access
Aug 2026
AraCTI-NER: A Dataset and Benchmark for Arabic Cyber Threat Intelligence Named Entity Recognition
A AraCTI-NER is introduced, a dataset of 10,312 token-level annotated samples over eight STIX-inspired entity types built by an LLM-assisted pipeline seeded with authentic Arabic cybersecurity articles, structurally validated and rebalanced through targeted generation.
Joud Alghamdi, Souham Meshoul
· Electronics · 0 citations