Rough-Set Based Association Rules toward Performance of High-Friction Road Markings

Abstract

The objective of this study was to explore the association rules influencing the frictional performance of high-friction road markings based on the skid resistance acceptance standard in excess of 65 British Pendulum Number (BPN). The paper integrates two widely used data mining approaches, rough set theory (RST) and the association rule algorithm, to extract the association rules. Three workshops with 14 experts and two on-site surveys were conducted, followed by collecting 303 field data sets based on British Pendulum Test method and ASTM E303-93 test standard. There are nine important attributes extracted and two of them are the core attributes: distance from margin and surface temperature. The 13 rules regarding the consequences of BPN <50 and BPN <55 indicate that surface age is vital in the skid resistance performance of road markings because it appears in every rule. There is more likely to be a decay in the skid resistance value as the surface age reaches the period between 11 and 15 months.