Mining Location and User Information from Users' Trajectories
MetadataShow full item record
The understanding of human mobility is integral to the advancement of many fields including public health, city planning, economic forecasting, and it attracts the interest of researchers from a broad number of fields. The work of such researchers would not have been possible without the availability of large localized datasets such as mobile phone Call Detail Records (CDRs) provided by telco operators and Location Based Social Networks (LBSNs) check-ins provided by popular applications such as Twitter, Facebook and Foursquare. However, such datasets are often incomplete, anonymized, and/or inaccurate. Hence there is a need to partially reconstruct or extract more information from such data. In this dissertation, we propose a framework to analyze and augment spatio-temporal data related to human trajectories that could be used to further the understanding of regularities in human mobility. In particular, the aim is to be able to augment the data to make it possible to classify users and places.