Web Knowledge Service Oriented Of Medical Information Classification Approach

From the Internet started to popularize , how to make full use of large, different structure, dynamic of Internet resources has become one of the core subject of information era. Information retrieval is for network users with the key technology of network knowledge service. But are also facing cannot provide good service and cannot make full use of network information application value challenge. Through processing of the large information , Network knowledge service provide real , crectly knowledge,not only the oriented information.Newwork knowledge service is the next development direction .There are four steps to provide knowledge service:knowdge acquisition,knowledge representation and knowledge validation and knowledge service.The current knowledge acquisition work is mainly theme search engine technology, but also faces a range of small, highly targeted challenge. In the knowledge representation, In 1998 ,Berners etc proposed the semantic web, but the semantic web itself did not resolve knowledge acquisition, reasoning and service, etc. This article study the network knowledge service, and puts forward the research oriented network knowledge service health text classification.The main research work have the following:.(1) Design and realized network knowledge service oriented health vertical search system, based on the vertical search engine crawler technique, classifier prediction methods affect crawler strategy, collect high correlation data.(2) Based on ontology semantic description for the knowledge representation of offers a generic, feasible method for reference. Using ontology construction technology, the paper constructs healthy body system, and based on this understanding, realize document document representation and knowledge reasoning.(3) Combined with healthy body system, established health knowledge base, for health text categorization provides the basis for support.(4) Through different feature selection method and different classification algorithm combined experimental, put forward based on improved TF - IDF feature selection algorithm of health classification algorithm, and verifies the proposed classification method in health classification superiority.This system has been published on the Internet. Experimental results and a user feedback has proved: this paper proposes the network knowledge service system structure, can better meet user requirements, improve the user experience degrees, reach the desired effect.