I’ll be hosting a Semantic Mediawiki workshop at MIT this upcoming Saturday and Sunday May 22/23 and invite you to attend. Semantic Mediawiki has made some significant progress on a topic I consider vital—helping end-users manage structured information. Semantic Mediawiki tackles this problem by adding a database to the Mediawiki platform (which runs Wikipedia) and letting people enter structured-data “facts” into the wiki pages they edit. The interaction is much the same as the current Wikipedia practice of editing infoboxes on wiki pages. The advantage is that once those infoboxes go into a database, you can query them. So, it is possible to embed queries in the wiki—e.g., give me a table of all presidents who were younger than 50 when they took office—-that produce tables of results from the database that show up when people visit wiki pages. Wikipedia is full of tables like this that are all edited by hand right now, which makes for much labor, data duplication, and opportunities for errors to creep in; a Semantic Mediawiki would construct those tables from the individual page infoboxes which would be much easier and more robust. Also, knowing about the structured data means you can help people fill in the infoboxes using forms that make it easier to fill in correct values. We’ve also experimented some with connecting our Exhibit framework to Semantic Mediawiki, creating an extension called Wibbit that lets you create rich visualizations of your data right inside wiki pages, much more easily than authoring a whole Exhibit by hand—you can see a bunch of examples of this on the Wibit site.
By leveraging the widely familiar wiki-editing workflow, Semantic Mediawiki creates an smooth path to incorporation of structured data in end-user-authored content. I think it has a great future, and look forward to seeing what progress we can make towards that future this weekend. I hope you can join us!