Research On The Chongqing Informatization Expert Consulting System Based On Ontology

With the development of information technology, more and more enterprises areeager to use information technology to enhance efficiency and improve the utilizationof information resources. Due to the complexity of the informatization, in the processof informatization construction，strong need information technology experts toconsult the project, to participate in the informatization project planning, feasibilitystudy, design project and plan selection. Therefore, whether to choose the appropriateexperts from the expert database is the key to successful implementation ofinformation construction for different types of information project. Traditional way ofexperts extraction used more random or based on keyword, Because both are toosimple, and lack of semantic, these methods can not guarantee the experts selected inthe project is the most appropriate. Therefore, taking the Chongqing informatizationexpert consulting system as object of the research, using the expression of ontologyin domain knowledge advantage, based on the information science ontology,researching experts and project matching from the perspective of semantic, can solvethe problem about experts extraction.This article studies the construction of informatization ontology, and designs thebuilding steps of informatization domain ontology. It extracts problem based on theconcept of the information domain ontology, participles and manages wordsegmentation and part-of-speech tagging on the documentation set. Filteringstop-words in the word segmentation result set by the stop words list gets candidateset of concepts. Against the shortage of the TFIDF on concept extraction, it proposesTFIDFI based on information entropy. Using this method, the concept of informationdomain ontology is proposed t from the candidate concepts. It completes theinformation technology domain ontology edit using protégé ontology editing tool.Then, on the basis of the information domain ontology, against the deficiencies oftraditional point-based and edge-based semantic similarity calculation method,andconsidering the depth of the concept in the information technology domain ontology,the density of the regions in which the concept of edge type, path between conceptsand the concept of the common attributes of semantic similarity, it proposes theimproved semantic similarity algorithm which is used to calculate the similarity ofthe project with experts. Finally, on the basis of studying the former two steps of Chongqing informatization expert consulting system architecture, function modules,database design, completes the development work of the system, and achieves thevarious functions of the system.Through two tests show the research results: First, compare the calculation resultof similarity algorithm proposed in this paper with artificial judgment, the fit degree ismore than80%; second, the average precision rate from the improved similarityalgorithm is44.1%, higher than the node-based algorithm32.53%and edge-basedalgorithm28.81%, and get the average recall rate which is51.5%,higher than thenode-based algorithm46.22%and39.66%of the edge-based algorithm. Theexperiments prove that the similarity algorithm which is proposed has a high fitdegree with artificial judgment, and can improve the precision rate and recall rate ofexpert extraction. Now Chongqing information expert consulting system has beendelivered to the Chongqing Economic and Information Technology Commission, andsuccessfully applies to Chongqing information expert consultation and reviews.