ISBN:  Unavailable

Category: Unavailable

Page count: 174

An empirical application using bike‐share system data shows that the method is useful for detecting different usage patterns of two heterogeneous user groups. The second study’s focus is to demonstrate our method in the similarity analysis for the semantic property of NTPs. The study uses the NTPs of individual bike trips to summarize the bikeability experienced by individual bike-sharing users. The proposed similarity measure for NTPs is utilized to assess the resemblance level of individual bikeability. Then, clustering analysis is conducted to generates the user typology of bike-sharing users based on individual bikeability. The third study develops and applies the FTP in the similarity analysis of movement trajectories' geographic context. The FTP is a generalized version of STP defined within velocity fields where the speed of moving objects is constrained by surface characteristics such as topography and land cover. The findings show that the similarity measures using the FTP-based methods are more robust to GPS uncertainty and produce more meaningful clustering results considering the geographic context of movement clusters than existing methods. The dissertation contributes to analytical time geography by developing similarity measures for NTPs and FTPs. The developed methods enable the quantitative comparison among different prisms, thus enhancing comparability of time geographic studies. In addition, the prism similarity measures support various data mining tasks from large movement databases to help find useful information about moving objects’ potential patterns affected by realistic constraints. Finally, this dissertation adds to the transdisciplinary turn toward integrated science of movement by introducing analytical tools for human and animal movement.