Had some good news today. One of our papers (with Jenny Liu
) on adaptive server architectures was accepted for publication at the 2006 Software Engineering for Middleware
(SEM 2006) workshop in Portland
in November. This is a fun, focused workshop that I’ve attended in the past and found very valuable.
This is the 4th version of this paper that we’ve submitted. We aimed big about a year ago and tried for ICSE 2006. Wildly ambitious, I know. With a 9% acceptance rate, this was always going to be a long shot. We got rejected, but two of the three reviewers were pretty positive, just one was not convinced. We took on board their comments, significantly upgraded the empirical validation, and submitted the new version in January to another leading conference, which shall remain nameless.
Of the three reviews this time, one was very positive, one neutral, and one absolutely hated the paper. Viciously. From the latter’s comments, it was however blatantly obvious this reviewer had absolutely no understanding of the paper whatsoever. I know this kind of thing happens, but this was a really bad case – probably the worst I’ve ever seen. We prepared a response for the Program Chair, but didn’t send it in the end. Program chairs have enough on their plate. We were unlucky, but you have to take these things on the chin in this game.
So we made some mods to improve clarity, and submitted to another major international conference in Melbourne. By this stage, we were very happy with the paper, and confident it’d be accepted. You can imagine our surprise when, of three reviewers, one liked it a lot, one was positive, and one hated it. This time we’d stumbled upon that classic reviewer profile of ‘this work refutes mine so it must be wrong’. Despite the strong empirical evidence in our paper, the reviewer dismissed our approach as infeasible. Never let the facts get in the way of entrenched opinions, I say.
Fortunately, the SEM 2006 submission deadline was timely, and our reviewers unanimously liked the paper. They even made insightful comments on how we could improve it. Luxury.
We know this is the first stage of a long R&D project, but it’s decent, innovative research and validated much more strongly than most current work in the adaptive/autonomic systems area. It’s good to finally find a qualified and informed audience. Ok, I’m biased
I am looking forward to Portland in November.