A Study on the Performance of CT-APRIORI and CT-PRO Algorithms usingCompressed Structures for Pattern Mining

Many algorithms have been proposed to improve the performance of mining frequent patterns from transaction databases. Pattern growth algorithms like FP-Growth based on the FP-tree are more efficient than candidate generation and test algorithms. In this paper, we propose a new data structure named Compressed FP-Tree (CFP-Tree) and an algorithm named CT-PRO that performs better than the current algorithms including FP-Growth and Apriori. The number of nodes in a CFP-Tree can be up to 50% less than in the corresponding FP-Tree. CT-PRO is empirically compared with FP-Growth and Apriori. CT-PRO is also extended for mining very large databases and its scalability evaluated experimentally.All these results point CT-PRO as the right candidate for generating a compact version of the original transaction database, which is small in size and which performs frequent pattern mining in a fast and efficient manner