An efficient mining algorithm for maximal weighted frequent patterns in transactional databases

In the field of data mining, there have been many studies on mining frequent patterns due to its broad applications in mining association rules, correlations, sequential patterns, constraint-based frequent patterns, graph patterns, emerging patterns, and many other data mining tasks. We present a new algorithm for mining maximal weighted frequent patterns from a transactional database. Our mining paradigm prunes unimportant patterns and reduces the size of the search space. However, maintaining the anti-monotone property without loss of information should be considered, and thus our algorithm prunes weighted infrequent patterns and uses a prefix-tree with weight-descending order. In comparison, a previous algorithm, MAFIA, exponentially scales to the longest pattern length. Our algorithm outperformed MAFIA in a thorough experimental analysis on real data. In addition, our algorithm is more efficient and scalable.