Open AccessA New Algorithm for Pattern Based Using Mining Association Rules
Author(s) -
S. Selvam
Publication year - 2020
Publication title -
asian journal of computer science and technology
Language(s) - English
Resource type - Journals
eISSN - 2583-7907
pISSN - 2249-0701
DOI - 10.51983/ajcst-2020.9.2.2171
Subject(s) - association rule learning , antecedent (behavioral psychology) , matching (statistics) , set (abstract data type) , point (geometry) , association (psychology) , computer science , institution , pattern matching , academic institution , data mining , algorithm , artificial intelligence , mathematics , statistics , psychology , developmental psychology , geometry , library science , political science , law , psychotherapist , programming language
It is indeed an art to match maximum number of preferences by utilizing limited number of resources. During the current academic year 75% of the admissions to Engineering Colleges have gone down, as only 30% to 40% of intake has been filled. Without reaching the breakeven point, the management of the institution becomes a complicated issue. The main aim of this paper is to discover a pattern to identify the choice of preferences of the candidates to seek admissions in any academic institutions. For the purpose of matching optimum number of candidates to suit our existing system, we have designed our algorithmic approach. Here our new system is used to extract frequent item sets from various preferences. By thresholds, it can fix the preferences either decrease or increase the level of frequent. The new algorithm is based on association rule classification which is one of data mining techniques. Here the frequency of itemset2 is combined with frequency to get itemset3 and continues until item set n. the new algorithm is easy to use and implement because its complexity is less. The application is designed to generate association rule until n-antecedent with one consequent. For this study purpose we have identified 15 most frequently used preferences among the students. The samples we have taken to get association rules are 100 students of Pannai College of Engineering and Technology at Sivagangai. The discovered pattern is common to all institutions. The pattern discovery may be accurate because it is computed by using factors like confidence and support. If this intelligent system is followed strictly, definitely the number of outcomes is increased. The applicant would prefer only when the supply is high. The result of this paper is an application that can generalize association rule among various academic institutions.