查看文章

With the explosion of the number of studies using big, passively-generated data for transportation analysis, this study focuses on understanding the properties of such data and how these properties affect our ability in deriving trip-related characteristics. Two big, passively solicited datasets were analyzed: a mobile phone data generated primarily on phone calls with locations identified through cellular triangulation and an app-based data generated primarily on app usage with locations identified through a mix of positioning technologies including GPS and cellular triangulation. Both datasets were compared against their household travel survey counterparts. It is shown that the two datasets, generated through different positioning technologies and usage mechanisms clearly have different spatial and temporal characteristics, which then affect trip related attributes such as trip rates and OD patterns. Implications in planning applications and future work are discussed.