If you think my enthusiasm for the DBpedia dataset and its possibilities is going overboard, just imagine the enthusiasm that's coursing through so many developers that there's something new on the scene so frequently. Their work certainly justifies my enthusiasm.

New one today is the DBpedia Relationship Browser, brought to you by Jens Lehmann and Jörg Schüppel at Agile Knowledge Engineering and Semantic Web (AKSW). Type in two objects, and it looks for a chain of relationships between them. It's based on the data provided in the infoboxes in Wikipedia. The infoboxes are interesting to me because they operate through, and unite, two very different kinds of structure. In doing so they are instructive, I think, about why the structured web is important.

The first kind of structure is structure that's useful to human readers--the visual structure. The bold heading, arranged basically a table structure, make up a structure that conveys a great deal of information visually. We know at a glance the types of information given. That's a testament to the rich history of writing -- after a couple thousand years of development, written text has got some pretty complex features. And, it's a testament to our fluency with written materials that we can know so much about a text literally at a glance, before we even get to actually reading it.

I should mention, too, that a recent conversation with Jeff McClurken brought up the subject of the choices made about what information goes into an infobox like this, and what that says about the person or culture producing it. Take this example, the English language infobox for Iraq. Do the choices about what data go into the box say something about how we understand nations and nationhood?

The second kind of structure is the structure that's useful for machines--especially RDF structure. DBpedia and the Relationship Browser makes particular use of the infoboxes, converting a structure that's useful for humans into a structure that's useful for machines, then turning that machine-useful structure into something that's once again useful for humans, but useful in an entirely new way. That's exciting stuff, and is one of the many things that make semantic/structured web projects resonate so well with the goals of higher education (analyze carefully, find patterns, structures, relationships, and use that to create new knowledge).

Even though it might not be quite up to such lofty ideals, I'll leave you with (one of) the inevitable searches with the tool, how to connect Kevin Bacon and Martin Scorsese.