Tag
This paper proposes MiNER, a fine-tuned BioBERT model for extracting biomedical entities from malaria-related clinical texts, and releases a human-labeled dataset for future research.
Introduces AutoSpecNER, an expert-annotated dataset for fine-grained named entity recognition in vehicle listings, with 659 advertisements annotated across 15 entity types. Benchmark results show DeBERTa achieves 90% micro-F1, outperforming rule-based and LLM approaches.
RedactionBench is a manually annotated benchmark for evaluating contextual PII redaction in large language models, introducing the R-Score metric and showing that contextual redaction remains an unsolved problem.
This paper introduces MHGraphBench, a knowledge-graph-grounded benchmark for evaluating large language models on mental health knowledge, including entity recognition, relation judgment, and multi-hop reasoning. Experiments across 15 LLMs reveal a gap between recognition and judgment capabilities.
This paper introduces IRC-Bench, a benchmark for recognizing implicit entities in first-person reminiscences using contextual cues rather than explicit mentions. It evaluates various LLM and retrieval configurations, finding QLoRA-adapted Llama 3.1 8B to be the top performer in open-world settings.