A Survey on Problems and Solutions of Frequent Pattern Mining with the use of Pre-Processing Techniques

Abstract

Frequent pattern mining is a researched area which is used for extracting interesting associations and correlations among item sets in transactional and relational database. Many algorithms of frequent pattern mining is been devised ranging from efficient and scalable algorithms in transactional database to numerous research frontiers and their wide applications. Many researches been done into FPM [1], but there are still several optimizations are required, so that FPM can be used more efficiently in data mining applications. For optimization purpose in many mining techniques data pre-processing plays an important role in reducing data size and also in lessening the time taken in database scans. This paper is a detailed study of problems and solutions of FPM techniques incorporated with pre-processing techniques. The intent of this paper is to summarize all major problems of FPM and their solutions. From this survey, it concludes that if FPM methods are merged with pre-processing techniques will produce results with better performance.