Identify the Problem

Analysts using the Savanna application often work with large data sets referencing people, places, and events. The relationships between any of those can be almost impossible to access if you’re just reading if on a spreadsheet. We already had the ability to display referenced places on a map, but we wanted to add visibility into temporal dimensions of the data, too.

Research

I interviewed analysts to better understand what kinds of questions they’re trying to answer with geo-temporal analysis, and to gather real-world use cases.  Though our users wouldn’t have been familiar with many of them, comparative analysis of tools that offered temporal visualization let me see what solutions had been tried.

I summarized my discoveries and began to clarify relevant user goals, then met with project managers and the development team to narrow the scope of work. We decided that initially we would support geo-temporal analysis, with the understanding that eventually we would expand beyond just map-based data.

Develop Concepts

At this point, I had a good idea of what goals we wanted to support, and I began sketching concepts about how to display time.

I explored a number of different ways to visualize time on a map and determined that a histogram that could also be used as a temporal filter would be the most useful place to start.

Concept review with stakeholders and user representatives helped keep the design on the right track.

Interaction and Interface Design

Once I settled on a general structure for the feature, I moved my work into Omnigraffle and began to lay out the interface and map the user's flow through it.  Lightweight prototype testing was used to refine the interaction sequences.

I was also able to identify enhancements for the next phase of the project.

Implementation and Outcome

I worked closely with the development team and our visual designer to bring the temporal filter into our map tool. A context-sensitive workflow lets map-makers easily add a temporal histogram to any map with dated points. The ability to filter the points significantly expands the usefulness of the map as a data exploration tool.