本文在计算学基因挖掘方法的基础上引入基因本体论GO（Gene Ontology， GO），从而提出一种融合GO的集成基因挖掘方法，即集成SDA-GO-SVM基因挖掘方法（Stepwise Discriminant Analysis，SDA；Support Vector Machine，SVM）。通过两套基因芯片表达谱数据实验并和其他方法对比，该方法能挖掘出更优的特征基因。

This paper proposes an ensemble-based gene mining method by incorporating GO（Gene Ontology，GO），namely an enseble gene mining method conbining SDA(Stepwise Discriminant Analysis，SDA)，GO and SVM(Support Vector Machine，SVM)．By validating the method on two public microarray datasets and by comparing with other gene mining methods，the results demonstrate that our method can mine better feature genes．