Font Size: a A A

Research On Key Technologies About Data Mining In CIMS Database

Posted on:2003-12-02Degree:DoctorType:Dissertation
Country:ChinaCandidate:D C PiFull Text:PDF
GTID:1118360092475981Subject:Mechanical and electrical engineering
Abstract/Summary:PDF Full Text Request
Data mining technologies have become one of the research hotspots today. In this paper the technologies about data warehouse, knowledge warehouse, active rule and knowledge management are studied at first. Based on the analysis the knowledge warehouse with activity and its architecture are put forward. This warehouse can make up for the data warehouse's lack. The active knowledge warehouse may become a research hotspot in the not too distant future.Secondly, about association rule, we put forward the TBAR* algorithm, the concept of tendency association rule and it's algorithm and filtering the rule based on person's belief. Based on the research of clustering analysis. A novel a clustering algorithm is posed, and the implementation methods of Prim and Kruskal are given too. The clustering algorithm of fuzzy ISODATA* is also raised based on the analysis of fuzzy ISODATA algorithm.Thirdly, an algorithm called multicommunity Genetic Algorithm is put forward to meet the application of attribute selection.At last, some problems about data mining standards, including mining language, architecture and mining process are discussed. An idea of integration of data warehouse with data mining is put forward. In the end of this paper a mining model with three logical architectures is posed. The algorithm storage with DLL and algorithm management with metabase can make the data mining platform with extensibility.
Keywords/Search Tags:Data Mining, Active Knowledge Warehouse, Tendency Association rule, Rule Filter, Rule Clustering, Multicommunity Genetic Algorithm, Algorithm management, Mining Platform
PDF Full Text Request
Related items