An Automated Association Rule Mining Technique With Cumulative Support Thresholds
C. S. Kanimozhi Selvi, Angamuthu Tamilarasi · 2009
Association rule mining is a task in data mining for discovering the hidden, interesting associations between items in the database. To find the relevant associations, the user has to specify support and confidence thresholds. These thresholds play an important role in deciding the number of appropriate rules found. User has many problems in specifying the appropriate thresholds, without the knowledge of itemsets and their frequency in the database. A high support threshold keeps away from generating more number of rules, but at the cost of losing interesting rules of low support. This paper proposes an approach to set suitable support thresholds for frequent itemset generation. Experimental results show that this approach produces the interesting rules without specifying the user specified support threshold.