Abstract:

Abstract enNumber of road motor vehicles that are handed over are increasing over the years. That makes the detailed examination of second hand car market is a necessity. It is possible to extract useful information about the operating mechanism of second hand car market using new analysis techniques and massive datasets. The purpose of this study is to examine the database which includes data belong to second hand cars by association rules. Association rules are successful at determining which two objects are observed simultaneously in a dataset. For this study, a dataset includes to 21109 second hand cars with 73 variables is scraped from a website in July 2016 and August 2016. In order to apply the apriori algorithm, the dataset must be consisted from Boolean variables. Numerical variables are converted to Boolean type by using their distribution. The association rules that are created with apriori algorithm is visualized with a network graph. At the end of the study rules such as “diesel cars have less fuel consumption, cannot go much faster, have higher torque values; new and expensive cars have higher taxes” are created. The overall results have useful information for those who operate in second hand car market.

Keywords: