رانندگان مسن تر و حوادث : تجزیه و تحلیل متا و کاربرد داده کاوی روی ازدحام دادههای حوادث

Teenage driving and associated accidents have been thoroughly studied. With the graying of our population in the United States, a focus on senior drivers and related accidents is needed. Unfortunately, there is not one comprehensive study that reviews the major existing studies conducted on senior drivers and accidents. In examining the literature, it also appears that data mining has rarely been applied in studying relationships between senior driver characteristics and accidents. This paper addresses these two needs by providing a meta-analysis of the existing literature on senior drivers and showing how data mining techniques could be used in this application.

The senior population is the fastest growing segment among the total population in the United States. From 1993 to 2003, the growth rate of the senior population increased by more than 15% of the total population (NHTSA, 2003). According to the 2001 NHTSA statistics, there were 19.9 million senior licensed drivers, which represented a 29% increase since 1992. In the United States, more than 40 million older adults will be licensed drivers by 2020 compared to 19.9 million older licensed drivers in 2002 as the baby boomer generation hits 65 years and older (Dellinger, Langlois, & Li, 2002). The number of senior drivers involved in police reported crashes is expected to increase by 178% by 2030. Drivers aged 65 and older will account for more than half of the total increase in fatal crashes and about 40% of the expected increase in all crash involvements. Senior drivers are expected to account for as much as 25% of total driver fatalities in 2030, compared with 14% presently (Lyman, Ferguson, Braver, & Williams, 2002). Older drivers now account for 1 in 6 accident fatalities, and as the elderly population grows, that number is expected to increase to 1 in 4 (Insurance Institute and Highway Safety, 2002). All these studies indicate that safety of senior drivers has become a challenging social problem in the US (National Highway Traffic Safety Administration, 2002 and Transportation Research Board, 1988). However, this also represents a social problem in other countries especially in Europe (Hakamies-Blomqvist & Peters, 2000) due to the increasing number of senior drivers, their high crash rate per mile driven, and their increased likelihood of injury. Accidents involving senior drivers should be diligently examined. The paper is organized as follows: Section 2 reviews the literature on senior driving patterns and characteristics; Section 3 analyzes the literature by comparing and contrasting the findings of the studies reviewed in the previous section; Section 4 is allocated to data mining applications of traffic accident data including a survey conducted on the Evergreen Society members in Montgomery County (MC), Maryland and its results, and the demonstration of how two data mining techniques may be used for mining traffic accident data to uncover hidden driving patterns; and Section 5 concludes the paper.