A Literature Survey on Association Rule Mining Algorithms

Abstract

With the dev velopment of d atabase techno ology, the need for data minin ng arises. As a r result, Associa tion Rule Min ning(ARM) ha s become a ve ery hot topic in d data mining. T This paper pres sents definition n and applicati on areas of ass sociation rules s. Furthermore e, a comprehe ensive literature review on th e existing algo orithms of ARM is conducte ed with a spec ial focus on the p performance a and application n areas of the al lgorithms. These algorithms ar re in general classified into th hree main class ses: (1) based on frequent item mset, (2) base ed on sequenti ial pattern, an nd (3) based on structured p attern. The algorithms are a e developed to improve t the accuracy and d decrease the complexity, an nd execution ti ime. However , it is hard to say y that they do always succee ed to optimize all these aspe cts simultaneous sly. Hence, there is still some space to s o develop more efficient alg rithms for different data stru o uctures.