I\'m using the feedparser library in python to retrieve news from a local newspaper (my intent is to do Natural Language Processing over this corpus) and would like to be ab
To expand on Bartek's answer: You could also start storing all of the entries in the feed that you've already seen, and build up your own historical archive of the feed's content. This would delay your ability to start using it as a corpus (because you'd have to do this for a month to build up a collection of a month's worth of entries), but you wouldn't be dependent on anyone else for the data.
I may be mistaken, but I'm pretty sure that's how Google Reader can go back in time: They have each feed's past entries stored somewhere.