Piggy Bank and more and more
Do you remember Piggy bank, a javascript-based very useful screenscrapper that extract data from web pages?
The same research team is now working in exhibits : an information visualizer in webpage form. This is a JSON serialized RDF mashup of information. This project try to achieve a very well known goal of the semantic comunity : make RFD applications in a transparent way. Something like: do RDF without doing it. The question is how will we get people to create RDF data sets in ways and for purposes meaningful to them?
May be is too early in our research to make this kind of questions.. but at the end, this problem will still be there… so.. he have to keep on mind user-friendly datasets creation when working with semantics.

