It sounds like you want to do a LawPivot, for general consumer application. You want to build a MARKET, which is the hardest thing imo to do in the startup world. Too many variables, and even with millions of dollars and a great founding team, you can fail (Airtime didn't fail, but for what they were hyped up as, I think it did). If you want to really do this, start with a small sector.
Btw, LawPivot would not work over tweets - it would ruin their professional presence (at least currently). Maybe you want to make a 411 tweet-bot of sorts? Where you tweet a question, and you answer it?
I feel as though you're hitting several different things at once, try to hit one thing, HARD.
Also, automate business to consumer, IN THE LONG TERM. Short term, anything goes. Mechanical Turk, you and a friend doing google searches after reading tweets, anything. Get paid, then automate. This is covered in "The Lean Startup", required reading.
That was just my stream of consciousness, thanks for reading!
Pinterest pulled a Steve Jobs - it was their job to know what consumer's wanted a year in the future. They were able to see that the coming smartphone explosion would combine with people wanting to be social, and what evokes a lot of emotion? Looking at old photos with friends. Add some filters on top to hide any bad picture taking skills/catering to the masses, and you have a billion dollar idea. They also provided a billion dollar execution, by exiting within two years. Pinterest was a case when all the stars align in the startup universe, and they also worked really really hard to build the telescope to see them.
What you can do: if the people are already on twitter, go to them. Create a twitter-bot that people can follow, and it automatically follows you back. When someone says "my back hurts", you use natural language processing on their tweets, and recommend a solution to them, gluing together Yelp, Angie's List, or anything else that might be relevant. Offer three months free, if they like it, they pay a monthly subscription. If not, take them off a list. When successful enough with consumers, you MAY be able to turn around and start charging businesses, but then, you have the responsibilities of curation, and I think it's easier to let companies who specialize in that ( like Yelp or Angie's List) do that, and you piggy back off them. Check if you're allowed to use their data, and if they charge for it.