sardonick

Late yesterday afternoon was an anti-climactic end to a weeklong binge of paper-writing, as my advisor and I sat down to examine our ICALT paper and realized that–because of lack of performance data on the ASR system that my work receives as input, and because of lack of overall pedagogical effectiveness of our project–the paper would likely be shot down if it was submitted right now.

It’s a bit frustrating because these are data that I have no control over getting. Part of shortcoming in doing research in a pipeline fashion, where one system feeds into another feeds into another–you are dependent on others for your results.

In retrospect (and, futurespect, as this is what I’ll need to be doing in the future so as to avoid this), I see that there are alternatives to only being able to look at my work in context of a pipelined system–I see I’ll need to build some experimental trials to isolate the effect of my system by swapping it out with a placebo system in some of the test subjects. The problem, of course, with when your software does something actually useful, people or institutions volunteer to be test subjects because they feel it will benefit them for the time they invest. In our case, they’re getting a free Arabic learning class. But it’s a hard sell to tell people “yes, we’re going to give half the people from your institution a dumbed-down version of our software, to measure effects. Nobody wants to be shortchanged. And, shoot, it’s demanding to build high-level tests for every part of the subsystem–we have enough trouble getting test subjects as it is, that it’s hard to spare enough people to make control groups for every subsystem we want to test.

It’s a bit of a let-down, not to be able to publish. It’s not that I’m not going to be able to get the information out ever (this conference was actually going to be the middle-step for a journal paper)… The real let-down is that I won’t be able to go to Kaohsiung for this conference. Was looking forward to being able to network with profs from China and Taiwan, as I’m considering future work in that area of the world as a post-doc or professor.