A Hybrid Approach for Biomedical Entity Name Recognition

Biomedical named entity recognition, an important step, makes preparation for extracting information from biomedical textual resources. This paper presents a hybrid approach to recognize biomedical entity, which includes POS (Part-of-Speech) tagging, rules-based and dictionary-based approach using biomedical ontology. Experiment results show our approach can find untagged biomedical entity name in the GENIA 3.02 corpus for aiding biologist tagging biomedical entity in the biomedical literature and obtain a recall of 66%, a precision of 78% and an F-score 71.5% for the test dataset extracted from the GENIA 3.02 corpus.