Performance of named entity recognition using exact matching for ME+BS in machine learning labeler and priority list in resolution module.

Each horizontal row shows a combination of features and the associated F-scores for each class on test data. ALL shows micro-averaged F-score. Key to external resources: J: JNLPBA model, U: UMLS and MetaMap, H: Human Phenotype Ontology, M: Mammalian Phenotype Ontology, G: Gene Dictionary from NCBI, L: Linnaeus, F: Foundation Model of Anatomy, P: Phenotypic Trait Ontology, C: Jochem's dictionary, B: Brenda Tissue Ontology.每一行展示了一组特征组合，以及测试数据集上各分类对应的F分数（F-score）。ALL列代表微平均F分数（micro-averaged F-score）。外部资源缩写对照表如下：J：JNLPBA模型（JNLPBA model）；U：UMLS与MetaMap；H：人类表型本体（Human Phenotype Ontology）；M：哺乳动物表型本体（Mammalian Phenotype Ontology）；G：美国国家生物技术信息中心（NCBI）基因词典（Gene Dictionary from NCBI）；L：林奈分类系统（Linnaeus）；F：解剖学基础模型（Foundation Model of Anatomy）；P：表型性状本体（Phenotypic Trait Ontology）；C：约赫姆词典（Jochem's dictionary）；B：布伦达组织本体（Brenda Tissue Ontology）。