Association Rule Mining through Matrix Manipulation using Transaction Patternbase

Abstract

In data mining studies, mining of frequent patterns in transaction databases has been a popular area of research. Many approaches are being used to solve the problem of discovering association rules among items in large databases. We also consider the same problem. We present a new approach for solving this problem that is fundamentally different from the known techniques. In this study, we propose a transactional patternbase where transactions with same pattern are added as their frequency is increased. Thus subsequent scanning requires only scanning this compact dataset which increases efficiency of the respective methods. We have implemented this technique by using two-dimensional matrix instead of using FP-Growth method, as used by most of the algorithms. Empirical evaluation shows that this technique outperforms the database approach, implemented with FP-Growth, in many situations and performs exceptionally well when the repetition of transaction patterns is higher. We have implemented it using Visual Basic which has substantially reduced coding and computational cost. Success of this method will open new directions.