On Developing Effectual Algorithms For Association Rule Discovery In Transaction Databases

Abstract

Association rule mining is the most important and well investigated data newlinemining technique used by an organizations decision makers to improve the newlineoverall profit It aids businesses to infer useful information on customer purchase newlinepatterns shelving criterion in retail chains stock trends and more Association newlinerule mining is usually carried out in two steps they are finding frequent item sets newlineand then using these item sets to identify the association rules It is well known newlinethat the first step frequent item set mining dominates the computational and I O newlinerequirements requiring repeated passes over the entire database As the volume of newlinedata in warehouses and on the internet is growing faster the scalability of mining newlinealgorithms is a major concern newlineClassical association rule mining algorithms that require more number of newlinepasses over the entire database can take hours or even days to execute and in the newlinefuture this problem will only become worse Sampling approach can be used to newlinesolve this scalability problem In the context of standard associationrule newlinemining use of samples can make mining studies feasible that were formerly newlineimpractical due to the enormous time requirements Indeed a number of large newlinecompanies routinely run mining algorithms on a sample of their data rather than newlineon the entire warehouse Especially if data comes as a stream flowing at a faster newlinerate sampling seems to be the only choice Current research efforts are focused newlineon inventing efficient ways of discovering these rules from large databases newlineLikewise most of the association rule mining algorithms presented in the newlineliterature are used for mining the static databases Nowadays research newlinecommunity has focused their researches into the incremental database on behalf newlineof the real world applications and the necessity of handling the dynamically newlineupdating new records newlineApplying data mining techniques to realworld applications is a newlinechallenging task because the databases are dynamic ie changes continuously newlinedue to addition deletion modification etc of the contained data newline newline