Open AccessFrequent Itemset Mining in a Unique Sc an using Transaction Database
Author(s) -
A. Subashini,
M. Karthikeyan
Publication year - 2020
Publication title -
international journal of innovative technology and exploring engineering
Language(s) - English
Resource type - Journals
ISSN - 2278-3075
DOI - 10.35940/ijitee.e2477.039520
Subject(s) - database transaction , scalability , data mining , computer science , apriori algorithm , benchmark (surveying) , database , gsp algorithm , a priori and a posteriori , association rule learning , feature (linguistics) , geography , philosophy , linguistics , geodesy , epistemology
In recent year, frequent Itemset Mining (FIM) has occurred as a vital role in data mining tasks. The search of FIM in a transactions data is discovered in this paper, pull out hidden pattern from transactions data. The main two limitation of the Apriori algorithm are undertaken, first, its scans the complete Databases at every passes to compute the supports of every itemset produced and secondly, the user defined responsive to variation of min_sup (minimum supports) thresholds. In this paper, proposed methodology called frequent Itemset Mining in unique Scan (FIMUS), needs a scan only one time of transaction databases to extract frequent itemsets. The generation of a static numbers of candidate Itemset is an exclusive feature, individually from the threshold of min_sup, which reduces the execution time for huge database. The proposed algorithm FIMUS is compared with Apriori algorithm using benchmark database for a dense databases. The experimental result confirms the scalability of FIMUS.