A new approach for efficiently mining frequent weighted utility patterns

Mining patterns that satisfy both frequency and utility constraints is an interesting problem in data mining. Currently, there are two main approaches to solving this problem: the first considers these two factors separately (skyline frequent-utility patterns), and the second combines these two factors to form a composite measure (frequent weighted utility patterns). In the second approach, the weighted utility support (wus), which represents the percentage occupancy of the total weighted utility of all transactions in a quantitative database, is used. This approach has the advantage of satisfying the downward-closure property, so it is easy to implement divide-and-conquer and pruning strategies on the candidate search space, as shown by algorithms such as MWIT-FWUI, MBiS-FWUI, and WUN-Miner. This study proposes an SWUN-list structure, a shortened version of the N-list structure, to effectively represent and mine frequent weighted utility patterns (FWUPs) in quantitative databases based on the second approach. Several theorems are also developed to quickly compute the wus values of patterns based on their SWUN-lists, as well as to determine the wus values of some special patterns without computation. The experiments are conducted on a variety of datasets to compare the proposed algorithm with the state-of-the-art methods. Experimental results confirm that the proposed method is superior to the existing methods for mining FWUPs in quantitative databases.

Automated summary

Mining patterns that satisfy both frequency and utility constraints is an important problem in data mining. There are currently two main approaches: one considers these factors separately, and the other combines them using a composite measure. This study proposes a new structure and algorithm to effectively mine frequent weighted utility patterns in quantitative databases, and the experimental results demonstrate its superiority over existing methods.