Social Web Information Retrieval for Crowds Mobility Management
Our cities need more efficient and dynamic planning. It has been estimated that by 2010 the cost of traffic congestion in the EU will reach 1% of the GDP. In 2007, congestion caused urban Americans to spend 4.2 billion additional hours on the road, which required an extra 2.8 billion gallons of fuel at the cost of $87.2 billion. These figures represent an increase of more than 50% in overall costs over the previous decade. Special events (strikes, gatherings, shows) represent only a small percentage of total traffic congestion but are responsible for very high and costly disruptions because they cause unexpected delays which neither travelers nor authorities are able to accurately predict.
Transportation planning has traditionally ignored such events focusing on average traffic impacts of land-use developments. Trip generation indicators are used for estimating trip ends for several types of venues, mainly large scale venues such as stadiums or conference centers. But it is impossible to predict event-specific impacts of those uses. Having an adolescent idol playing at a concert hall is not the same as having an orchestra: the number and behavior of the attendees changes radically. InfoCrowds explores the interactions between online information about public events, mobility data and event specific surveys to build explanatory and predictive models of flows of people, and their transportation mode, in the city.