Font Size: a A A

Research And Realization Of Digital Plans Management System Based On Text Clustering

Posted on:2014-03-27Degree:MasterType:Thesis
Country:ChinaCandidate:Q LinFull Text:PDF
GTID:2268330425981305Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
At present, our country is in a critical period of economic transition. Strictly control ofthe incidence of all kinds of accident and improvement of the ability to prevention andrespo nse to emergency, are ine vitable de mand of good and steady economic growth andexchange ruggedly type growth mode. And this is also a significant aspect of building aharmonious society and testing the functions of the government’s performance. Faced withfrequent types of emergencies, how to make continge ncy plans maximize effective ness is thefocus. Most the plans are stored in the form of paper and low informatization in our country atpresent. Moreover, the plans have wide variety. It is hard to provide the function ofinformation filtering and integration. And it is not conducive to emergency rescue work.Therefore, this paper put forward a conception of digital emergency plan managementsystem. Aiming at the disadvantage of present plan management system, artificial intelligenceand data mining are combined with research content of provincial-emergency platform, whichis to build and a reasonable and effective plan management system and provide intelligentservice for political appointees.In regard to problems abo ve-ment ioned, thorough analysis and research are carried on inthis paper. Firstly, the structured storage prob lem is solved by the frame theory of artificialintelligence. And then cor relation technique of digital plan mode ling and knowledgerepresentation of plans are introduced. And plan matching method based on the text similarityis caught up with, business process of matching is designed, Similarity of inversion lists isgenerated. DBSCAN algorithm of text clustering is improved. The original Eps is initializedby data fitting. And adjustable neighborhood Eps is adopted in the iterative process. Thismethod reasonably adjust plan clustering discriminate conditions, strengthen the intensitybetween data of clusters, solve the problem of boundary point identification mistake to someextent and improve the efficiency of the algorithm. Finally, the topic met the requirements onbusiness of plans management, designed system architecture, function modules, and databasesto complete the research and realization of digital emergency plans management system.
Keywords/Search Tags:digital plans, frame theory, similarity calculation, text clustering, DBSCANalgorithm
PDF Full Text Request
Related items