ABSTRACT

Mining frequent patterns has always been a great field of research for investigators. Various algorithms were developed for finding out frequent patterns in an efficient manner. But the major drawback of all these researches is the increased number of database scans. Partition algorithm is one of the approaches for mining frequent patterns but the large number of database scans required in this algorithm makes the mining process slow. Few developments have succeeded in reducing the number of database scans to two. Here an attempt has been made to develop a K-Partition algorithm which requires one database scan. Whole database is compressed in the form of Karnaugh Map, having very small size i.e. a fraction of the whole database. Then partition algorithm can be used to identify frequent patterns using K-Map model. Thus this approach brings efficiency in terms of time taken by processor for mining frequent patterns.