Penentuan pola sekuensial data transaksi pembelian menggunakan algoritme SPADE

Abstract

The rapid development of information technology that is happening these days requires people to adapt into these developments. This human efforts can be seen from the many activities carried out by computerization that produces large amounts of data. From these available abundant data, the discovery of useful knowledge from large database becomes popular and attractive. This discovery of useful knowledge can be done using the concept of sequential pattern mining. One of the algorithms that applies the concept of sequential pattern mining is Sequential Pattern Discovery using Equivalence classes (SPADE) that is able to determine the sequential pattern of a data transaction. By adopting the functions contained in the SPADE algorithm, the purchasing tendency of items by customer at a specific time period can be seen. This research use the purchase transactions data of Sinar Mart Swalayan in period of 1 March to 31 March 2004. In this research, the minimum support was tested starting from 45% to 89% and minimum confidence from 20% to 96%. Minimum support and minimum confidence which is given is determined based on the condition of the data. Experimental results showed that the maximum value of minimum support that still could generate frequent sequences was 89%.