Title :
Cross Correlation based Inter-Transaction Association Rule Mining Technique
Author :
Ghanem, A.M. ; Tawfik, B. ; Owis, M.I.
Author_Institution :
Fac. of Inf. Syst., Suez Canal Univ., Ismailia
Abstract :
Several algorithms have been proposed to solve the problem of mining frequent inter-transaction item set. However, the low efficiency of support calculation for inter-transaction itemsets is still a challenging problem that eliminates the performance of mining algorithms. This paper provides inter-transaction association rule mining algorithm using effective technique for support calculations. The proposed technique is based on cross correlation and bitwise operations. The experimental results show a significant improvement of performance up to several orders of magnitude compared to First Intra Then Inter (FITI) algorithm.
Keywords :
data mining; database management systems; knowledge based systems; medical computing; transaction processing; FITI algorithm comparison; First Intra Then Inter comparison; bitwise operations; cross correlation; intertransaction association rule mining technique; mining algorithm performance; support calculations; Association rules; Biomedical engineering; Data mining; Electronic mail; Information systems; Irrigation; Itemsets; Terminology; Transaction databases; Cross Correlation; Data Mining; Inter-Transaction Association Rule Mining;
Conference_Titel :
Biomedical Engineering Conference, 2008. CIBEC 2008. Cairo International
Conference_Location :
Cairo
Print_ISBN :
978-1-4244-2694-2
Electronic_ISBN :
978-1-4244-2695-9
DOI :
10.1109/CIBEC.2008.4786036