There is a growing library of visualization tools being developed (eg on GitHub) - for Stanford's Data-Driven Documents. The implementation languages and standards are the usual W3C suspects: JS, HMTL5, CSS, SVG, JSON etc. Here's some links and examples.
Bostock, M., V. Ogievetsky, and J. Heer, D(3): Data-Driven Documents. IEEE Trans Vis Comput Graph. 17(12): p. 2301-9.
- Stanford Viz Group. 2012 http://vis.stanford.edu/.
- Bubble Chart. 2012 http://mbostock.github.com/d3/ex/bubble.html.
- Dendogram. 2012 http://mbostock.github.com/d3/ex/cluster.html.
- Forced-Directed Graphs. 2012 http://mbostock.github.com/d3/ex/force.html.
- Population Pyramid. 2012 : http://mbostock.github.com/d3/ex/population.html.
- Crossfitters. 2012 http://square.github.com/crossfilter/.
What do MMA users feel is the best strategy for interfacing to this platform , given that MMA and WebMMA are analytic systems but are positioning to overlap the market segment for web-based data visualization.
Likely because MMA is proprietary, the community of MMA users and developers is much smaller than open-source systems, so that visualization libraries built on D3 (but also also include systems like cytoscape.org &c) will have a higher growth rate. Also WRI is only ~500 employees so there's a limit to implementing new functionalities.
My Q is what is the path of least friction to interface MMA analysis code to systems like D3? How to carve the turkey at its joints? This is a soft Q, not expecting code examples, but more comment-driven discussion.