MEMORY EFFICIENT FREQUENT PATTERN MINING USING TRANSPOSITION OF DATABASE

Frequent pattern can be extract from any dataset by using Apriori algorithm. Apriori algorithm is first choice of all researchers to find frequently occurs pattern from any binary dataset. Dataset contain record of user purchase item as transaction record. This paper improves existing apriori algorithm performance by extract frequent patterns from binary transaction data. New approach is applied for dataset implementation in form of transposed database of user’s record for fast data access. New work has done to mine frequent patterns using transposition of dataset, if database is large and contains thousands of attributes but having only some objects. This work analytically investigates the search space problem of frequent patterns mining and characterize database in transposed form and proposes an algorithm for mining frequent patterns based on Apriori algorithm by space reduced longest common sequence (LCS).This method makes apriori algorithm space efficient. The space complexity of proposed algorithm is O(n) while the dynamic approach like longest common subsequence space complexity is O(n2) memory for given items in dataset.