Communications of the ACM

How to understand evaluation criteria for CS researchers.

Anonymous

If the seed venues are "central" or "important" to the area, then instead of including JCB and TCBB for Bioinformatics, the authors should have considered journals like Genome Research, Genome Biology, PLoS Computational Biology, PLoS Biology, Cell, Nature, and Science -- these journals by far have a higher impact overall and publish high quality research in Computational Biology which I can not assuredly say for JCB and TCBB.

Jacques Wainer

The goal of the seed venues for an area is to be sure that if someone co-authored a paper in that venue, it is very likely that the researcher works in the area, and is accepted as so. Given that, we could not include journals such as Nature, Cell, Science since one cannot say that anyone that published in Nature works in Bioinformatics.

Second, we used the data on the list of authors available at DBLP as the universe of researcher from which to sample. But Genome Research and Genome Biology are not listed on DBLP.

Thus, from the commenter's list, only PLoS Computational Biology could have been used as a seed, and indeed it is a possible alternative. The important question is whether the set of researchers that publishes in PLoS CB is sufficiently different than the set of researchers that publishes in either JCB or TCBB that the universe of Bioinformatics researchers would have been very different, which may change some of the results in the paper. We believe it is not.

-- the authors