The Mining Ontology-based Research On The Relationship Of Scientific Network

With the widespread popularity of the computer in various fields and the rapiddevelopment of Internet, the amount of information was an explosion of exponentialgrowth. The 21st century, the global amount of information is doubling the rate ofincrease every three years. So，in order to respond the challenges that brought by theinformation explosion, we urgently need some automation technology to help peoplequickly find information they really need from huge information. InformationExtraction is a way to solve this problem.Information extraction technology is extracting a specified event or factualinformation from a natural language text, and describes information with thestructured form. Its ultimate aim is to develop useful information extraction system,extracts the information which users interested in from free text. The research of thispaper is using Ontology technology to help complete the information extraction workin information extraction system. The introduction of ontology is the guarantee theconsistency of the structured and the data. On the basis of previous work, this papermainly to complete the following job:(1) Introduced the background and significance of this study, and analyzes of thecurrent research status of the Ontology and the Information Extraction.(2)Study the theories ontology-based information extraction system, mainlyincluding ontology and information extraction technology. Discuss the ontologydefinition, classification, modeling primitives, building rules and methods, languageand ontology building tools. Second, study information extraction in-depth, includingthe concept of information extraction, information extraction systems, andontology-based information extraction system.(3) On the basis of analyzing the structural features of the scientific collaborationnetwork, proposed one kind method to build the domain ontology oriented scientificcollaboration networks, and use this method to build the target object-oriented domain ontology of scientific collaboration networks. Using protégé4.1 defined the conceptof domain ontology, its data attributes and object attributes, and use the protégéownreasoning tool parsing and reasoning the domain ontology (containing concepts,relationships, constraints, and many other information describing the field) , ensurethe accuracy and consistency of the ontology. Using ontology parsing can also extractdomain ontology information contained in the field, such as the concept of ontology,the relationship between the concepts, the relationship between the domain and range,instance of the class, and so on.(4) To unstructured text use natural language processing technology，andcomplete pre-processing and segmentation operation，then produce the pretreatmentof document sets. Design of the frame structure of the ontology-based informationextraction system, and analyze the important modules.(5) According to the result of ontology parsing generates information extractionrules, use this rule do extract information extraction operations on the pretreatmentdocument generated. And the information extraction results will save in accordancewith the demand of structuring scientific collaboration networks. Finally, using socialnetwork visualization tool Ucinet to finish scientific collaboration networkvisualization and analysis.