Visualization of dynamic socio-economic geographies of cities
City maps are the standard expressions to convey a sense of geographical order into the urban landscape. When looking at them, there is always the impression that things are neat and precise: buildings, roads, and administrative boundaries are depicted according to their coordinates and (usually) in the most objective manner as possible. But how much do they reflect the varying perspective of city dwellers? Indeed, cities thought as living entities present a wide variety of interactions among their main actors: people share places, benefit from services, give functionalities to the architecture, create interactions and connections among different areas of the urban fabric, define personal geographies going beyond official borders. The whole urban net is made up by complex systems of knots coming from both social and economic structures reflecting how people actually perceive city spaces. Understanding and visualizing such dynamics are major challenges to contribute to effective urban policies being able to ensure / support / promote / enhance competitiveness, sustainability and social cohesion and prevent / minimize negative effects in urban environments. Thus, successful approaches to the urban problem should take into account alternative representations of the city dynamics - not captured in alternative approaches - in order to achieve a complementary but powerful insight of the real urban space. The specific goal of this contribution is to understand connections among city neighbourhoods by mapping their affinities against specific variables of interest. It is implemented by means of highlighting and representing patterns in people movements across the urban context due to both economical (e.g. shopping and home-work journeys) and social (e.g. migration or geographical distribution of friends? network) reasons. Our analysis attempts at depicting dynamic geographies out of people social and economic activities, highlighting citizens? behaviours across both time and city-space dimensions, and from here contributing to foster decision-making policies as well as urban simulations. Association discovery rules are the main techniques we are using in order to reach this aim (see Section II). By measuring people activities across given city sites, it should be possible to visually highlight places sharing a common likeness, whose identities represent a new way to look at geographic aggregation (see Section III). Indeed affinities and similarities will provide new city clusters standing for a closer representation of the real appearance of a city for its inhabitants. As opposite to classical, administrative area definitions, activity boundaries are inherently overlapping, inter-playing, domain-dependent and changing over time.