Tisa Mitprachya. Frequent Itemset Mining by Using Length of Transaction and Sum of Elements. Master's Degree(Computer Engineering). King Mongkut's University Technology Thonburi. Library. : King Mongkut's University of Technology Thonburi, 2004.
Frequent Itemset Mining by Using Length of Transaction and Sum of Elements
Abstract:
Finding regularities in the shopping behavior of customers of supermarkets is anexample application of data mining. The essential of this application is finding frequentitemsets which occur in database. The main problem of the approach for findingfrequent itemsets is scanning database several times therefore it takes much time andtoo large of memory. The recent algorithm that is faster than the former algorithm isvertical mining algorithm, because it supports fast frequency counting via intersectionoperations on transaction ids. To reduce the mining process time and used memory sizeduring finding frequent itemsets mining process, this research proposes a verticalmining algorithm about finding frequent itemsets mining by using length of transactionand sum of elements. We apply our algorithm to the Eclat program which is one of thetwo best-known basic algorithms for mining frequent itemsets in a set of transaction.Generally, vertical mining algorithms check the support of transaction list after doingintersection. Length information that we use in preprocessing pruning is the newapproach in vertical mining algorithm. If length information of current data in processmatches with our length information, it will be pruned. With this reason, our algorithmcan achieve reducing processing time.
King Mongkut's University Technology Thonburi. Library