Conferences vs. Journals in CS, what to do? Evolutionary ways forward and the ICLP/TPLP Model

Abstract

We computer scientists seem to do it differently to other sciences: we publish mostly in conferences and our conferences are of a different nature. Our journal papers are long and take a long time to review and publish whereas often their papers are short and published quickly. And all this interacts with the tendency to evaluate researchers or departments frequently and in a mechanical way (via paper numbers and citation counts) instead of infrequently and deeply (by actually reading papers) and the fact that the current way in which bibliometry is done makes our papers invisible to the world. This position paper offers my viewpoint on what the problems are and why they are important, and also elaborates on some realistic ways forward. In particular, regarding the issue of conferences vs. journals, it proposes the model adopted a few years back by the logic programming community for its main conference (ICLP) and journal (TPLP, Cambridge U. Press). This model is based on the assumption that CS journal papers can be of two types: rapid publication papers (similar to those of other sciences and also close to our conference papers) as well as the longer journal papers that are traditional in CS. Then, the concrete proposal is to, instead of publishing the traditional conference proceedings, have the papers submitted instead to a special issue of an (indexed) journal which is ready online in time for the conference. The traditional conference reviewing process is also improved (following journal standards for rapid publication papers and special issues) to include at least two full rounds of refereeing and a final copy editing step. I argue that this model offers an evolutionary path that solves a good number of the incompatibility problems with other sciences of the current CS models, without giving up the essence of CS conferences. For this reason I believe that this model (or one of the closely related models being proposed) would be a clear path forward, and relatively straightforward for the community to adopt widely.