An efficient algorithm for mining top-rank-K frequent patterns from uncertain databases

The analysis and management of uncertain data has gained a lot of importance in the past few years because of their importance in a wide variety of applications such as sensor network and privacy preserving data mining applications. Many algorithms have been proposed to mine the frequent pattern over uncertain database. However the existing algorithms for uncertain data generate a large no. of candidate patterns and required to define an appropriate user defined threshold which is a challenging task for users. In this paper, we propose a new algorithm known as UFAE (uncertain filtering and extending) algorithm to mine top-rank-k frequent itemset or patterns. Mining only top-rank-k frequent pattern greatly decrease the number of candidate pattern generated so reduce the mining time. Many algorithms exist to mine top-rank-k frequent itemset in case of precise data but none in case of uncertain database. Experiments are performed to evaluate the performance of the algorithm on various dataset.