Title :
A hybrid k-DCI and Apriori algorithm for mining frequent itemsets
Author :
Suriya, S. ; Shantharajah, S.P.
Author_Institution :
Dept. of Comput. Sci. & Eng., Coll. of Eng. & Technol., Madurai, India
Abstract :
Data mining involves analysis, extraction, refining and representation of required data from large databases. kDCI (k Direct Count and Intersect) algorithm is one of the best scalable algorithms for identifying frequent items in huge repository of data. This algorithm uses a special kind of compressed data structure which helps in mining the datasets easily. Apriori algorithm, a realization of frequent pattern matching, is universally adopted for reliable mining. It is based on parameters namely support and confidence. kDCI algorithm is hybridized with Apriori algorithm for better performance than their individual contribution. The result proves scalability and mining speed effectively.
Keywords :
data compression; data mining; data structures; apriori algorithm; compressed data structure; data mining; frequent itemsets mining; frequent pattern matching; hybrid k-DCI algorithm; k direct count and intersect; mining speed; scalable algorithms; Data mining; Itemsets; Signal to noise ratio; Weight measurement; Apriori; Confidence; Scalability; Support; kDCI;
Conference_Titel :
Circuits, Power and Computing Technologies (ICCPCT), 2013 International Conference on
Conference_Location :
Nagercoil
Print_ISBN :
978-1-4673-4921-5
DOI :
10.1109/ICCPCT.2013.6529038