DOI: 10.5176/2251-2195_CSEIT17.63
Authors: Asst. Prof. Muna S Al-Razgan
Abstract: Association rule mining reveals interesting associations or relationships among items of a large dataset. Association rule initially used for a market basket data set; market basket analysis is an important technique in the retail industry. However, most research has been conducted using artificial rather than real data sets because it was not possible to access current real life market basket data sets. This study presents the process of collecting, preprocessing, initial assessment, and association rules of a real life Arabic market basket dataset as part of a course assignment for undergraduate students in a IT43 course. Initial evaluation and analyses of the Arabic market data indicated that the preferred purchasing date and time for customers are Saturday evening and that they usually pay by cash. Moreover, the most bought items were chocolate and cheese.
Keywords: data mining; association rules; FP-Growth Algorithm; market basket analysis; Arabic dataset
Updating...