Identifying named entities in biomedical text based on stacked generalization

Biomedical named entity recognition is a basic technique in the biomedical knowledge discovery and its performance has direct effects on further discovery and processing in biomedical texts. In this paper, we present stacked generalization strategy for biomedical named entity recognition including homogeneous classifier ensembles and heterogeneous classifier ensembles based on stacked generalization. Evaluations show that stacked generalization strategy can take advantage of more useful evidences, and make use of compensation and relativity among different classifiers to learn the correlation between individual classifiers predictions and the correct prediction to improve the performances of the system. This method breaks through the limitation of single classifier and achieves promising performances.