Current many information retrieval methods are based on purely keywords for representing the user needs. One of the main problems with this method is that it does not formally capture the explicit meaning of a keywords query, and ignores the documents that may be different in content but related with them. To improve the precision of the information retrieval, this paper proposes a hybrid method for information retrieval in digital libraries, and gives a corresponding framework and the algorithm of the approach. The proposed method combining the ontology technique and vector space model can improve effectively the drawback existed in present retrieval system.