It makes me proud that's at my university. :)   The English department people mostly have put together the texts, while a few ethnomusicologist friends of mine have been in charge of getting recorded samples. It is such a huge undertaking, that they have gone about the very practical approach of having "regular people" just sing them. There is SO much material, I believe the goal is just to get some sort of recorded rendition in existence, so that ballad scholars might have tune matched with text. To this end, classes on broadside ballads were taught, and the students in those were recruited to record. Other interested parties were recruited, too. I recorded around 16 of them, maybe, though I must admit they were pretty poor -- the time issue, you see, and not being able to really study the material beforehand. Finally, the directors of that aspect of the project (whom I won't name here, but who will be credited on the site) generously gave their time to pound out better renditions of many many ballads.

To give an idea of the quick 'n' dirty process (in my case, at least), I was given a set of broadsides on all of which was indicated that they were sung to such 'n' such air. The melody of the air was given separately. It was challenging trying to match the verse to the melody; sometimes it didn't work very well. Pronunciation was a big issue, too, but one that we could not really spend time on.