Visualising Data with Simile

The Evaluation Group at CARET has been looking at and experimenting with MIT's Simile toolkit, as one (relatively user-friendly) way to use semantic tools for data representation and exploration. We are increasingly interested in thinking about incorporating semantic technologies into "social computing" approaches and tools.

A big advantage of Simile is that it hides a lot of the heavy-duty programming (in this case, Java) from the user, enabling easy / streamlined building of interactive web pages. A disadvantage is that the more we use it (at least until we get our own version installed!), the slower it gets, because whenever a local user loads a locally-hosted page with Simile stuff on it, the page then has to make a call on the MIT server.

These kinds of data visualisation tools are developing rapidly, but I think it's still early days yet and we are still very much novice-apprentices. I'm very much aware that there is a level of sophistication available to information scientists and design experts that we haven't gotten anywhere near yet. I'm scanning sites like FlowingData, Visual Complexity, and Many Eyes, which have a lot of nice examples of how to visualise data for effective communication and analysis. Junk Charts keeps it real, with analysis of how charts and visualisation can be used misleadingly, to fudge data that doesn't fit. Perhaps most excitingly, their samples don't all rely on the Google Maps API...because why should Google Maps be our only visual interface to map data? Projects like MySociety.org are leading the way, making (largely open-source) contributions to social well-being and community action, with neat map/web-based tools to make your life better, like their excellent travel-time maps.