Cities have adopted evolving urban digitization strategies, and most of those increasingly focus on data, especially in the field of public transportation. Transportation data have intuitively spatial and temporal characteristics, for they are often described with when and where the trips occur. Since a trip is often described with many attributes, the transportation data can be presented with a tensor, a container which can house data in $N$-dimensions. Unlike a traditional data frame, which only has column variables, tensor is intuitively more straightforward to explore spatio-temporal data-sets, which makes those attributes more easily interpreted. However, it requires unique techniques to extract useful and relatively correct information in attributes highly correlated with each other. This work presents a mixed model consisting of tensor decomposition combined with seasonal vector autoregression in time to find latent patterns within historical taxi data classified by types of taxis, pick-up and drop-off times of services in NYC, so that it can help predict the place and time where taxis are demanded. We validated the proposed approach using the experiment evaluation with real NYC tax data. The proposed method shows the best prediction among alternative models without geographical inference, and captures the daily patterns of taxi demands for business and entertainment needs.

Services provided by public facilities are essential to people's lives and are closely associated with human mobility. Traditionally, public facility access characteristics, such as accessibility, equity issues and service areas, are investigated mainly based on static data (census data, travel surveys and particular records, such as medical records). Currently, the advent of big data offers an unprecedented opportunity to obtain large-scale human mobility data, which can be used to study the characteristics of public facilities from the spatial interaction perspective. Intuitively, spatial interaction characteristics and service areas of different types and sizes of public facilities are different, but how different remains an open question, so we, in turn, examine this question. Based on spatial interaction, we classify public facilities and explore the differences in facilities. In the research, based on spatial interaction extracted from taxi data, we introduce an unsupervised classification method to classify 78 hospitals in 6 districts of Beijing, and the results better reflect the type of hospital. The findings are of great significance for optimizing the spatial configuration of medical facilities or other types of public facilities, allocating public resources reasonably and relieving traffic pressure.

This master thesis report presents the development of a geographical visualization system using taxi data. The system uses a large data base from a taxi company that have previously never used the data for visualization purposes. The taxi company requested a system that processes the data on a server on demand and visualizes it on a web client using heat map visualization as a primary visualization technique. The web client was supposed to be easy to use, provide deeper knowledge about the business of a taxi company and at the same time kept interactive with low latency for data requests. A big part of the thesis focuses on techniques for decimating an original data set to a smaller representational data set to be used for heat map visualization and sent to a web client from a server. The project continues by optimizing the system to keep latency to a minimum and finally developing a web client to explore the data. The result is a system with promising latency that is easy to use for exploring data and gaining a deeper knowledge about a taxi business.