Biomedical named entity recognition (NER) using second-order the conditional random fields model

Abstract

—Biomedical named entity recognition (NER) is an important technique for biological information extraction. Here, we have proposed a model using machine learning where biological terms are recognized using second-order conditional random fields (CRFs). From our experiments with the JNLPBA 2004 dataset, the CRFs labeling model returns an F-score prediction value of almost 99%.