Maintenance of Discovered High Average-Utility Itemsets in Dynamic Databases

• Published in: Applied sciences Vol. 8; no. 5; p. 769
• Main Authors: Zhang, Binbin, Lin, Jerry, Shao, Yinan, Fournier-Viger, Philippe, Djenouri, Youcef
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 11.05.2018
• Subjects: Algorithms; Data mining; Decision making; International conferences; Knowledge discovery; Maintenance; Science; Upper bounds; Denmark; Vietnam; Norway; China
• ISSN: 2076-3417; 2076-3417
• DOI: 10.3390/app8050769

[...]retailers or managers cannot take efficient decisions or choose appropriate sales strategies on the basis of the most profitable sets of products. In this case, some rules may arise after the database update. Since the FUP concept applies an Apriori-like approach, which may be inefficient, Hong et al. developed the FUFP-tree (Fast Updated Frequent Pattern tree) algorithm to update frequent itemsets for transaction insertion [26]. An itemset X is a high average-utility upper bound itemset (HAUUBI) in the original database if it satisfies that condition: HAUUBID←{X|auub(X)D≥TUD×δ} For instance, assume that the minimum average-utility threshold is set to 14.4%. [...]the itemset (a) is a high average utility upper bound itemset (HAUUBI) since auub(a)D (= 68 > 244 × 14.4%)(= 35.13). [...]this sorting order of itemsets generally does not change much when the database is updated. Since all 1-items of inserted transactions are kept to build the initial AUL-structures, the designed FUP-HAUIMI algorithm can efficiently find the updated HAUUBIs and HAUIs without generating candidates and is complete and correct to update HAUUBIs and HAUIs.