WSEAS Transactions on Computers
Print ISSN: 1109-2750, E-ISSN: 2224-2872
Volume 15, 2016
Probabilistic Space-Time Analysis of Human Mobility Patterns
Author:
Abstract: Human mobility models are widely required in many academic and industrial fields. Due to the spread of portable devices with positioning functionality, such as smartphones, ordinary people can obtain their current position or record mobility history. Thus, mobility history can be processed in order to identify human mobility patterns. The human mobility pattern can be analysed in two ways: space and time. Space analysis focuses on a users location, and time analysis emphasises a users mobility on a daily basis. From the raw positioning data of various sets of user mobility, we analysed a personal human mobility model. Each users positioning data set is pre-processed and clustered by space and time. For spatial clustering, we developed a mechanism of clustering with expectation maximisation methodology. For temporal clustering, stay or transition probabilities over a 24 hour period were analysed. We represented the result of the personal human mobility model using the continuous time Markov chain (CTMC) with spatial transition probabilities over time. We developed a process to construct a personal human mobility model from a persons positioning data set. This personal mobility model can act as a basis for many other academic or industrial applications.
Search Articles
Keywords: Human Mobility Model, Space-Time Analysis, Location Clustering, Continuous Time Markov Chain, Individual Personal Mobility
Pages: 213-229
WSEAS Transactions on Computers, ISSN / E-ISSN: 1109-2750 / 2224-2872, Volume 15, 2016, Art. #21