Abstract—Mining association rules is one of the most
important and popular task in data mining. Current researches
focus on discovering frequent itemsets that is an important step
to it. Many algorithms for discovering frequent itemsets have
been proposed. However, for a large database, an efficient
mining algorithm must be a better balance in I/O cost and main
memory load. Most traditional algorithms, like Aprioir
[Agrawal, 1993], often take higher I/O cost because of
multi-scan over the analyzed database. There have been a few of
algorithms, like FP-Tree [Han, 2000], use a limited pass
numbers to databases, but they could suffer from the shortage
of main memory as there does not consider time constraints to
association rules. In the paper, we first discuss the problem of
mining temporal association rules in databases. Then, we create
the necessary sub-operators between itemsets and interval
operators between time intervals to mine temporal association
rules. Finally, a new algorithm called MTAR_Sub for mining
temporal association rules is designed and discussed.
Index Terms—Association rule, data mining, frequent
itemset, network traffic, temporal constraint.
Guojun Mao is with the School of Information, Central University of
Finance & Economics, Beijing, China, 100081 (e-mail:
maximmao@hotmail.com).
[PDF]
Cite:Guojun Mao, "Mining Temporal Association Rules in Network Traffic
Data," International Journal of Future Computer and Communication vol. 2, no. 6, pp. 55-59, 2014.