Deriving information from structured data

May 15, 2008

On net@nite I heard that Dipity, a great site for mashing up timelines uses Freebase as one of its information sources. An excellent example is this timeline for Marillion embedded below

Earlier today, on Friendfeed, Alex Iskold described an “aha” moment for Powerset, which I have been quite underwhelmed by thus far (although it has a gorgeous UI). When I browsed over to the search link, it seemed to be that the reason the results appeared as they did were because they came from Freebase, whose structure made the query very powerful. Wonder if others think that’s the primary reason as well? The Wikipedia results for the query were not that impressive.

So what do these results tell us? Well, perhaps there is value in taking data and adding structure the way Freebase and dbpedia are doing. The ability to generate different representations, etc is quite powerful. I can already envision what Pierre might do with some of these tools and the work he has been doing with Freebase and Wikipedia.

These examples are somewhat trivial, but it is not difficult to imagine more interesting scenarios, perhaps to represent protein interactions, or trying to find non-obvious relationships and linkages. Will be interesting to see what kinds of applications are developed on the backs of these data platforms

Technorati Tags: Powerset, Freebase, Wikipedia, Semantic Web, Structured Data