ISI and Conferences

ISI from Thompson Scientific might be seen as just another scientific article indexing service, just like Google Scholar, Citeseer and many others. But it has a much stronger effect: many universities only “count” ISI-indexed publications. In mainstream operations research, this doesn’t have a very strong effect. Most well-known OR journals are ISI-indexed, and those that are not strive greatly for such indexing. But in some of our newer fields, particularly those related to computer science, this is much more of an issue. In constraint programming, conferences are key outlets for research, particularly such conferences as CP and CP/AI-OR. Both of these conference publish their papers in Springer’s Lecture Notes in Computer Science, which up until recently was ISI-indexed.

The ISI people have recently dropped LNCS and other series from their main ISI-indexing, and created a “conference” version of ISI-indexing. For some, this can have a very strong effect on promotion and tenure.

I have somewhat mixed feelings about this. At one level, I am worried that the ISI-“page count” available for some of our fields has decreased greatly. This will make it harder for some of our subfields to grow and expand. On the other hand, conference volumes are just part of the proliferation of research outlets, and not ISI-indexing them seems appropriate given the sometimes quick level of review that some conferences provide. It is certainly true that the conferences provide reasonably thorough reviews and the rejection rate is high (70% or more rejected), but this still seems different than journals. Perhaps it is just me: as a member of a conference committee, I feel it is my responsibility to review papers somewhat farther from my core interests than I would for journal reviews. As such, I can’t provide the same insights for conferences that I do for journals.

Much of OR doesn’t have this problem. Most OR conferences are “free-for-alls” with no veneer of reviewing. This makes for large conferences of highly-variable quality. It is great for talking about not-fully-formed ideas, or interesting but not publishable results. I wonder if some of the competitive conferences will change in flavor now that the “carrot” of an ISI-indexed paper is not offered. At the very least, people should be working harder to follow up the conference with a journal publication, which seems a good outcome.