Reclamation of extraordinary utility items in the multi-level catalogue

Utility mining is the central area of data mining that reveals a data set with high-use items (HUIs). To date, several techniques to extract HUIs in a single component database have been projected. According to the literature survey, multilevel data sets do not have utility mining algorithms. HUIs of a multilevel data set is the primary goal of this work. This paper proposes to find HUIs in a multilevel dataset with a unique utility mining algorithm called MUMA (Multilevel Utility Mining Algorithm). To supply the valuable data of both the item set, MUMA utilizes a parse tree known as MUMT, a multi-layered efficiency mining tree. This tree produces candidate item sets that reduce the method’s space complexity. The research has also suggested a reinforced tree-based algorithm called MU UP Mining, Multilateral Utility Mining, MU UP, and Lexicographic Tree (MU LG). The study also proposed an enriched tree-based algorithm. These experiments are done using various datasets such as transaction datasets, weblog data, and synthetic datasets, with all the suggested algorithms. Each dataset evaluated performance factors such as performance time, reminiscence space, the number of potentially high utilities, and the number of HUIs collected on each level.