Dataviz and infoviz

Useful summary of the status quo for data visualization by futurist Melanie Swan (via Eloise Pasteur's blog about her new open source histogram tool funded by Melanie).

As someone who is aiming to take StoryMachine into this area (well, infoviz maybe more than dataviz), I'm very sympathetic to the concept. However, the reality can be a little different on a limited prim, 30 metre parcel with all the constraints built into LSL to avoid griefing. Of course, that doesn't mean we shouldn't try.

My wish-list for StoryMachine continues to grow faster than my capacity to code. I previously mooted browse and authoring modes in addition to playback but I now also want to implement a search mode as well.

In particular, I intend implementing a PubMed search (PubMed being the online implementation of the Medline database that draws together metadata and abstracts of papers published in medicine and related areas). The NCBI provides a set of eUtils that make it fairly straightforward to lash together a RESTful type URL to encode the search terms.

Of course, SL doesn't help too much in terms of handling text and data (it may ultimately be necesssary to hand-off some management to the Karrigell server) but it should be fairly easy to rez sets of prims to represent searches and hits and to play around with positioning and particles to add value. Whether it could be a viable search tool is questionable but the fact that a teacher or librarian might be present to prompt and guide adds a useful dimension. This made me wonder whether it hadn't already been done: in fact, it has on HealthInfoIsland but apparently via Google and limited to 5 hits which leaves scope for refinement.

This increased functionality proposed for StoryMachine raises the issue of whether the modes or result sets might usefully interact. And, of course, whether there is scope for dataviz too (currently this is limited to the Google Charts API).