I scienced all last weekend. Except for a few hours when my fiancée and I watched The Hobbit on DVD, I scienced almost nonstop, and I actually have something to show for it.
I… can’t actually say what it is, because I’m putting together a conference paper based on this research, and part of the terms of submission is that the research not appear anywhere previously in any form, so while I can’t even say what I’ve been researching yet, I can say that so far my hypothesis seems to be borne out in the results.
As with any kind of empirical experiment, this is pretty exciting—and all the more so for me because what I’m doing is, as far as I know, quite knew, and there’s not much research on the topic at all. I’m still in the middle of gathering data, but based on a preliminary set that I believe is pretty representative of what I expect to gather by the end of this whole excruciating process, it looks like I was right.
Now comes the hard part, of (first finishing the data gathering and then) tweaking exactly which features I want to extract from it and how I want to weight them to get the most consistent and correct results. And then I have to finish writing both the conference paper and the thesis itself.