Efficient high-utility itemset mining over variety of databases: A survey

Abstract

High-utility itemset mining (HUIM) is creating new heights of challenge in the areas of research in big data analytics blending data structures, AI, and machine learning techniques to find efficient utility patterns. In the recent past, the merchandise market researchers were pretty much interested in analyzing the purchasing patterns and forecast new profitable area of business. HUIM is that one kind where we get business intelligence and has become the most basic method offinding the knowledge in decision making and optimizing the decisions of market analysis, streaming analysis, biomedicine, mobile computing, stock exchanges, etc. HUIM was initially applied widely in the transactional databases, later applied recursively in incremental databases, further moved to dynamic databases like temporal, spatial, and data stream databases. This chapter highlights the insights of various existing algorithms present in the literature based on several applications and tries to update the various availabilities of different approaches for coining latest research in this domain.