Research And Implement Of Chinese Medical Ontology Retrieval Based On Jena

Recently, most of information retrieval systems are based on traditional keywords match technology with which computer can not understand information meaning. It causes result of retrieval content with a lot useless information and professional literature information retrieval facing challenge. Ontology retrieval is based on concepts and relationships, which change the way of retrieval from keywords match to content match. Compared with traditional keywords match, it have great advantage in retrieval result. In the field of Chinese medical literature retrieval, the result of retrieval is unsatisfactory, which cause problem to the professor of Chinese medicine.According to the demand of Liaoning Chinese medicine university professor and the Chinese Medicine coronary disease Ontology file they provided, importing the ideal of Ontology retrieval to the project, taking Eclipse as programming tools, using Java as programming language, importing Jena Package, creating Ontology in RDF Model, retrieving information with SPARQL language, the Chinese medical literature retrieval system was designed after researching. Because of present information extraction process in Excel file unable to structural information, an information extraction method was proposed considering the structural feature of Chinese Medicine coronary disease Ontology file. The method can fetch data information and structure information at the same time, which meet the needs of creating knowledge library. According to the result of information extraction, analyzing resource and property that Chinese Medicine coronary disease involved with, knowledge library was created which makes computer know concept of Chinese Medicine coronary disease terms and relationship between them. This article show the method of retrieving information in knowledge library with SPARQL language. The method needs two query template. The use of two template can implement retrieving term keywords and relationship between them. The RDF model was stored at file system by the method Jena provided which improve operating efficiency. Recently, most of the semantic relevancy of concept calculation methods are based on shortest path, which do not consider global semantic distance and relationship proportion impacting on result. A synthesize semantic relevancy of concept calculation method was designed which has consider global semantic distance and relationship proportion impacting on result. The contrast experiment show that synthesize semantic relevancy of concept calculation method is more suitable for Chinese Medicine coronary disease Ontology.The Chinese medical literature retrieval system implements: searching keywords related to the keywords inputted; searching prescription recorded by literature related to inputted keywords; searching symptom that cored by prescription related to inputted keywords; searching the closest relationship between two keywords. After testing the system, it shows that: the Chinese medical literature retrieval system meet the needs of Liaoning Chinese medicine university professors, Ontology retrieval technology makes computer know the concept of Chinese medical terms which solves the weaknesses of traditional keyword matching technology and raise both recall ratio and precision ratio.