The Future of Accessible Graphs

A core component of an accessible website is the provision of alternative descriptions of non-textual content — like photographs — via such HTML structures as the ALT and LONGDESC attributes. And a lot of the time, a simple sentence of ALT text outlining the content of the image is basically OK. But what do you do when you’re required to publish complex graphs online?

Currently, the only way to produce a graph accessibly is to put in the elbow grease and match an image of the graph with a LONGDESC summarising the data that underlies the graph. With a bit of luck, you might have access to the original spreadsheet that generated the graph and you can then think about providing the information in the LONGDESC (which can be a separate HTML document) as an HTML data table.

The maturation of such formats as Scalable Vector Graphics (SVG) and even the currently non-standard Canvas extension suggest a way to automate the work of providing a textual alternative. Put simply, if we can generate complex data graphics (such as graphs) from a data-set on the fly with javascript, the textual version of the data must be already available somewhere. It’s just a different transform to tablular data — instead of SVG markup.

Is it happening already? Anyone who wants to post examples of this kind of thing is more than welcome.