An Analysis on Various Measurements For Association Pattern Identification In a Multiple Database

Exploring Objective Measures and Their Applications in Association Analysis

Abstract

Data mining isan area of data analysis that has arisen in response to new data analysischallenges, such as those posed by massive data sets or non-traditional typesof data. Association analysis, which seeks to find patterns that describe therelationships of attributes (variables) in a binary data set, is an area ofdata mining that has created a unique set of data analysis tools and conceptsthat have been widely employed in business and science. The objective measuresused to evaluate the interestingness of association patterns are a key aspectof association analysis. Indeed, different objective measures define differentassociation patterns with different properties and applications. This paperfirst provides a general discussion of objective measures for assessing theinterestingness of association patterns. It then focuses on one of thesemeasures, h-confidence, which is appropriate for binary data sets with skeweddistributions. The usefulness of h-confidence and the association pattern thatit defines—a hyper clique—is illustrated by an application that involvesfinding functional modules from protein complex data.