Big Changes Already

On Monday I learned that my immediate supervisor, and the only person in the group who has been devoting their entire time to the question-answering project, is going on at least a 3-month leave. She’ll be gone starting the first week of September.

A newly-minted Ph.D. will be starting the first week of October, and while he might contribute somewhat to the QA project, he has his own things to do, in addition to getting acclimated to BBN.

One of the other guys in the group is starting grad school part-time, and it sounds like he’ll be working on the same project as me, but for the school instead of for BBN. I’m not sure how that works in terms of IP, but such is the nature of government research programs.

The public information about AQUAINT isn’t very interesting. It’s basically just a research directive. I’m not exactly sure who runs ARDA (Department of Fisheries and Hatcheries, I bet), but I think they’re the ones paying for most of me (see my recent post on billability) for at least the next several months.

Of course, this all adds up to me having a fairly important role in the direction and implementation of the non-linguistic front-end and mid-end of the system. I guess I’m ready; my main worry is that I’ll make some change that someone won’t like, or will break some weird case that I didn’t know to test for.

I think my first priority is to develop test harnesses for all of the various modules, and really improve the error handling and explicit exception throwing so that errors don’t cause the mid-end and back-end modules to just disappear without returning any results.

The Show Must Go On” from Greatest Hits III by Queen & Elton John