Biomedical Named Entity Recognition Based on Classifiers Ensemble

In this paper, we present classifiers ensemble approaches for biomedical named entity recognition. Generalized Winnow, Conditional Random Fields, Support Vector Machine, and Maximum Entropy are combined through three different strategies. We demonstrate the effectiveness of classifiers ensemble strategies and compare its performances with standalone classifier systems. In the experiments on the JNLPBA 2004 evaluation data, our best system achieves an F-score of 77.57%, which is better than most stateof the art systems. The experiment show that our proposed classifiers ensemble method especially the stacking method can lead to significant improvement in performances of biomedical named entity recognition.

• Type: Journal
• Year: 2008
• Where: IJCSA
• Authors: Haochang Wang, Tiejun Zhao, Hongye Tan, Shu Zhang
• Keywords: Conditional Random Fields, Entity Recognition, IJCSA 2008, Support Vector Machine