Hi Otis,

Haven't had a chance to experiment with this yet. Its on my list of experiments once we get our new test server set up. I think the hard part will be finding good test cases with our data. I'm also interested in trying to determine if we have to tweak the default length normalization because we have such long documents. I plan on looking at SweetSpot Similarity. We really haven't started looking at evaluating/tweaking relevance ranking yet.

Tom