VLDB09 - journal like review process.

This year's vldb was held in Lyon, France. I came back from there yesterday night. The jet-lag woke me at 5pm...so I have some time to do some notes here.

The first thing i want to note is the undergoing change of the review process of Proc. of VLDB. This year' s vldb includes a couple of (<=10) papers that have gone through a journal like review process(multiple rounds), rather than the traditional one round thing for conference. Next year's proceeding will include 10% of such papers. From 2011, there will be no traditional conference PC and everything will be shifted to the journal-like track.

I talked to quite a few of my friends. Most of them did not seem quite enthusiastic about the transition. I guess the reason might be that they are able to get papers accepted anyway and the multi-round process only implies an increasesing workload but not a higher chance to them. Literally quote (translated into english) some of their words "I bet the review will be 'more experiments on xxx', 'more references on xxx' , ' more xxx on xxx' "; "more experiments, on a new dataset, these are only easy to say"," the worst thing could happen is that a reviewer didn't like the paper intially but asked for a major revision rather than gave a outright reject. Then the paper was turned around multiple times, the content was rewritten and exp was redone and the reviewer still didn't like the paper and finally rejected it"......I have a few journal papers and went through the multiround process myself twice. Fortunately, my experience was not that bad. The first paper was submitted to transaction on information theory. we got two quite positive reviews. One review was like 5 pages long with numerous detailed comments, but each of those was fairly easy to fix and the paper got accepted after that round. The other one was a very short paper which i submitted to Information Processing Letter. The process was even smoother. I got two reviews and one of them only had one sentence for recommending acceptence. Responsing the other review took me less than two hours and the paper was accetped after this round. I hope this new vldb journal thing can improve the paper quality substantially without bringing us too much pain....