A FP-Growth based method for extraction of sequential patterns from quantitative databases

The problem of discovering sequential patterns is to find inter-transaction patterns such that the presence of a set of items is followed by another item in the time-stamp ordered transaction set. So far, many methods have been proposed to discover the sequential patterns. One of the effective approaches within them is to extract the patterns by using FP-Growth. However, FP-Growth approach contains important disadvantages; including the use of complex data structures and the requirement of some recursive processes to extract sub-trees. Moreover, in the past, many algorithms were proposed for mining sequential patterns, most of which were based on items with binary value. Transactions with quantitative values are, however, commonly seen in real-world applications.In this thesis, in order to overcome the problems mentioned above, first, using item-based candidate generation instead of recursive processes, a FP-growth based novel approach is proposed. This method is then adapted to find sequential pattern in databases with quantitative item. The experimental results conducted on real data set of Dicle University, Faculty of Medicine, Center Laboratory show the applicability and the superiority of the proposed method.