Data characterization by means of novel spatio-temporal patterns

Bike Sharing Systems are sustainable transportation strategies able to reduce the greenhouse gas emission. In recent years, they registered a strong growth thanks to the various benefits they bring, such as solving the "first and last mile" problem, that consists in travelling for short distances to reach work and get back. This thesis analyzes a data-set containing data from Barcelona's stations to extract some meaningful patterns from a specif position and timestamp into a discretized space and time. The research consists in conducting different analyses with the Prefix Span algorithm to search for some correlations to predict future events that could happen. The procedure used in this research can be applicable to other data-sets that contain events that happen over time, for which the spatial information is known.