Dissertation > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer software > Program design,software engineering > Programming > Database theory and systems

Mining and Evaluating Bridging Rules

Author ChenFeng
Tutor ZhangShiChao
School Guangxi Normal University
Course Computer Software and Theory
Keywords outlier rough set entropy bridging rule
CLC TP311.13
Type Master's thesis
Year 2006
Downloads 42
Quotes 0
Download Dissertation

With dramatic development of technology, the ability of producing and collecting data has also been improved constantly. On the contrary, that of dealing with data is not enhanced at the same ratio. It seems for people to analyze so a great deal of data, which causes an embarrassing phenomenon that“drowning in data but starving for knowledge”. Consequently, data mining emerged not only for acquiring useful knowledge from data but also for analyzing and finding out valuable information.Data mining[1] is a process of nontrivial extraction of implicit, previous, unknown and potentially useful knowledge from a large amount, incomplete of noisy, fuzzy and random data. The knowledge extracted is represented by such fashions as concepts, rules, regularities, patterns.Data mining associates closely with many other fields including artificial intelligence, pattern recognizing and machine learning and so forth. Now it includes several important branches which are association rules, classification, clustering, time serial analysis and outlier (exception pattern) analysis.There always exists data objects that do not comply with the general behavior or model of the data. Such data objects, which are grossly different from or inconsistent with the remaining set of data, are called outliers[24-26]. Some existing algorithms in machine learning and data mining have considered exceptions, but as noises, and excluded them out of analysis. Indeed, from the point of knowledge discovery, rare events are often more interesting and valuable than others. Examples of its applications include the detection of credit card fraud, the monitoring of criminal activities in electronic commerce and many other areas[27-34]. Therefore, exception pattern mining is an important research work.Some typical algorithms for detecting outliers include statistical based[35], distance based[36], deviation based[37-38], depth based[39] and density based[40]. They all defined and mined outliers from different points of view, and got remarkable progress.Firstly, this paper introduced the existing data mining techniques and the disadvantages and advantages of current outlier mining algorithms. Then, a new definition of outlier is put forward: bridging rule. This points such rules that the antecedent and action of which belong to different conceptual clusters. And it represents some interaction between conceptual clusters. For example, the antecedent and action of“diaper→beer”belong respectively to alcohol and baby goods, so it is an bridging rule. Such outliers have been widely used, for some representation of

Related Dissertations
More Dissertations