An Efficient Data Mining Algorithm Based On Item Sets Frequency and Priority Using Distribution Model

Abstract

ABSTRACT newlineFrequent pattern mining casts a vital part in many of significant data mining tasks such newlineas associations, sequential patterns, partial periodicity, to name a few. Nevertheless, it is newlinecommon knowledge that frequent pattern mining habitually produces an extreme number of newlinefrequent item sets and rules, paving the way for reducing in competence as well as efficiency newlineof extraction in view of the fact that clients are best with the task of sieving through a huge newlinenumber of extracted rules to locate the fruitful ones. Therefore, without resorting to the newlineextraction of the frequent itemsets, mining only closed frequent itemsets goes a long way in newlineincredibly enhancing the excellence along with the decrease in the computational time. newlineMany methods have been proposed for mining items or patterns from data base. These newlinemethods use frequency for extracting patterns from the data base. But frequency based newlineextraction is not always successful. In addition, frequency methods have some drawbacks. newlineTo overcome these drawbacks, the utility (priority) based method was introduced. Utility newlinebased methods extract patterns or items based on the weight or priority of the items. The newlineindividual performance of these methods over the history of data base mining has drawbacks. newlineAccordingly, many works were developed using both frequency and utility methods and newlinesuch works perform satisfactorily in mining items from the data base. But, these works do newlinenot provide assurance that the extracted patterns will continue to provide the same level of newlineprofit and frequency in the future. No literature work is available to solve this drawback. newlineTo overcome this problem, a mining algorithm was proposed for extracting patterns from newlinedata base using both frequency and utility methods. But this method has the drawback of newlinememory usage and processing time. Because, in data streams data elements arrive at a rapid newlinerate. The incoming data is unbounded and probably infinite. Due to high speed and large newlineamount of incoming data, frequent itemset mining algorithm must re