Urban Data Analytics
Using probabilistic formalism, we model the macro-level processes observed in a city (commuting, shopping, tourism, etc.) as temporally varying mixture distributions over atomic space-time urban actions conditioned on physical city infrastructures. The class of models we develop is based on learning probabilistic decompositions of the co-occurrence tensors containing disaggregated data of individual actions and interactions of city dwellers.
We take a conceptual view on the system through the set of individual time-stamped and geo-referenced events of interactions within and between a bipartite network of citizens (a representation of a social network) and a bipartite network of venues in a city (a representation of city infrastructures).
Availability of data opens new perspectives on activity-based modelling of urban mobility and traffic micro-simulations.
- McArdle G., Furey E., Lawlor A., Pozdnoukhov A., City-scale Traffic Simulation From Digital Footprints, UrbComp’12 at ACM SIGKDD, 2012
- Kaiser C., Pozdnoukhov A., Enabling Real-time City Sensing with Kernel Stream Oracles and MapReduce, Pervasive and Mobile Computing, Volume 7, Issue 5, pp 708-721. 2013
- Kling F., Pozdnoukhov A., When a City Tells a Story: Urban Topic Analysis, In ACM proceedings of the 20th ACM SIGSPATIAL GIS, 2012
- Lawlor A., Coffey C., McGrath R., Pozdnoukhov A., Stratification structure of urban habitats, Pervasive Urban Apps at PERVASIVE’2012.