Human activity spatio-temporal indicators using mobile phone data

In the context of Smart Cities, monitoring the dynamic of the presence of people is a crucial aspect for the well-being of an urban area. We use mobile phone data as a proxy for the total number of people (Carpita & Simonetto 2014), with the specific aim of computing spatio-temporal region specific in- dicators. Telecom Italia Mobile (TIM), which is the largest operator in Italy, thanks to a research agreement with the Statistical Office of the Municipality of Brescia, provided to us about two years (April 2014 to June 2016) of High- Frequency Daily Mobile Phone Density Profiles (DMPDPs) in the form of a regular grid polygon each 15 minutes. Densities have to be rescaled in order to express the total amount of people rather than just TIM users. Separately for selected regions in the province of Brescia, characterized by being either working or residential areas, we group similar DMPDPs and we characterize groups by their spatial and temporal components. In doing so, we propose a mixed-approach procedure. First, borrowing the method of the Histogram of Oriented Gradients (HOG, Tomasi 2012), we perform a reduction of the DM- PDPs dimensionality computing their features extractions. With this method, we convert a 2D spatial object into a 1D vector of data, by preserving the spa- tial relationship contained in the data. Secondly, we stack in a single vector all the HOG features of the same day and, by applying a high-dimensional cluster analysis that accounts for the curse of dimensionality, we group days. Third, for each group, we reshape the data in order to form a 3D array with dimen- sion a (quarters), b (days) and c (space), and we apply a Canonycal Polyadic (CP) tensor decomposition (CANDECOMP/PARAFAC, Kolda & Bader 2009) to extract three indicators related to the dynamic of the presences along the space, the days and the quarters.