PII Masking Benchmark for Icelandic (MIM-GOLD-NER, augmented)

PII Masking Benchmark for Icelandic (MIM-GOLD-NER, augmented) A training/benchmark set for Icelandic PII detection, built by taking all 3 splits of vesteinn/icelandic-ner-MIM-GOLD-NER (train, validation, test — same sentences, same split boundaries) and layering PII-specific annotations on top of its existing named-entity gold labels. It was created to benchmark privacy_filter, a rule-based Icelandic PII masking tool, but the annotations are generic and usable for any Icelandic… See the full description on the dataset page: huggingface.co/datasets/Sigurdur/pii-masking-mim-gold.