Research and Design on Personalized DL Based on J2EE

From the angle of natural language processing, this paper analyses the architecture and core technologies of personalized digital library. It introduces matching strategy of document content representation and topic search. This matching strategy is a expansion model of vector space model. It uses word relating matrix base on tradition VSM. The WRM is computed by co-occurrence and mutual information of index word. Experiment data show the strategy used WRM is better than tradition VSM. At the same time personalized digital library is designed based on J2EE according to above strategy. It provides a project for personalized DL searching information in isomer database. This paper also introduces the key technologies based on J2EE, which includes EJB delegate technology, joint common pool technology and so on.