Taxi poco

Visually explore large spatial-temporal city data: Study on New York City Taxi Trips

Summary: As more and more municipal information is collected and made available, there are new possibilities for data-driven analyses that can improve citizens' living through evidence-based decision-making and policy making. This article focuses on a particularly important dataset: taxi rides. Taxi cabs are precious detectors and information associated with taxi driving can give unparalleled insights into many different facets of metropolitan living, from business to personal behaviour to pattern of travel.

However, the analysis of this information poses many challengingities. It is a highly sophisticated set of information containing geographic and chronological elements and several associated variable values for each journey. It is therefore difficult to specify explorative searches and carry out comparable analysis (e.g. comparing different areas over time). In New York, there are an estimated 500,000 taxi rides per night on a daily basis.

Our proposal is a new type of taxi that allows the user to request taxi rides manually. In addition to analysing standards, the scheme also provides support for origin-target surveys, which make it possible to investigate urban transport throughout the town. Our results show that this paradigm is capable of expressing a broad spectrum of spatio-temporal interrogations, and it is also agile, since not only can interrogations be written, but also various kinds of aggregate and display visuals can be used, enabling the user to research and benchmark results.

We developed a scaleable system that implemented this paradigm that supported interactivity reaction time, used an Adaptive Level-of-Detail-Rendering strategy to create a clear visualisation for large results, and showed the user invisible detail in a synopsis using hot map Overlays. Featuring a number of case histories inspired by transport and economics professionals, we show how our system and system enables domains professionals to accomplish functions that were previously impossible for them.

More than half of the world's inhabitants live in metropolitan areas for the first consecutive year. Whereas in the recent past decision-makers and sociologists have been confronted with significant limitations in gathering the necessary information to grasp metropolitan dynamism and assess policy and practice, today the information is plentiful. A number of towns have begun to provide a broad set of datasets, see e.g.[24],[10],[7].

However, the problem is how to use this information effectively.