A Novel Approach for Finding Representative Pattern Sets

Frequent pattern mining produces large number of frequent patterns. So it creates challenge on visualizing, understanding and analysis of generated patterns. So there is need of finding a small number of representative patterns without affecting understanding and analysis of generated patterns. Few algorithms have been reported in the literature that satisfies these needs. However, these algorithms result in either higher time or space complexities and also do not scale well with the size of frequent patterns. An approach that overcomes the problem of time and space complexity has been reported in the literature that allows the user to select the tradeoff between space and time. However, with large size of frequent patterns, an incremental approach should perform in a better way. This paper deals with the development and analysis of an efficient incremental algorithm for finding minimum representative pattern sets.