Using the Pre-large Concept for Maintaining High Fuzzy Utility Itemsets

Utility mining is a popular research field in data mining. It uses item utilities and quantities to fit more into real applications. Fuzzy utility mining has also been proposed to reflect the linguistics of human perceptron for item association. In the past, we proposed a fast-update-based (FUP-based) approach to maintain high fuzzy utility itemsets for continuously coming transaction data. This paper proposes an algorithm that applies the pre-large strategy on fuzzy utility mining to raise the maintenance performance. Nine cases are considered to maintain the current high fuzzy utility itemsets based on the fuzzy upper-bound utility. The results of the numerical experiments show that our proposed algorithm has better efficiency than the batch and the FUP-based approach in the execution time.