The explosion in the volume of data about urban environments has opened up opportunities to inform both policy and administration and thereby help governments improve the lives of their citizens, increase the efficiency of public services, and reduce the environmental harms of development.

However, cities are complex systems and exploring the data they generate is challenging. The interaction between the various components in a city creates complex dynamics where interesting facts occur at multiple scales, requiring users to inspect a large number of data slices, over time and space.

Conventional manual exploration of these slices is ineffective, time consuming, and in many cases impractical. In this work, we propose a technique that supports event-guided exploration of large, spatio-temporal urban data.  We model the data as time-varying scalar functions and use computational topology to identify potential events in different data slices.

To handle a potentially large number of  events, we develop an efficient and scalable algorithm to group and index them. This allows users to interactively explore and efficiently query event patterns on the fly.

A carefully designed visual exploration interface helps guide users towards data slices that display interesting events and trends. We demonstrate the effectiveness of our technique on two different data sets from New York City: data about taxi trips and subway service. We also report on the feedback we received from analysts at different NYC agencies.


Moore Sloan Poster Research Project