Database Systems (TODS)


Search TODS
enter search term and/or author name

Social Media

June 2015 ACM TODS Editorial

Editorial: The Best of Two Worlds: Present Your TODS Paper at SIGMOD?

It just became even more attractive to publish your research results in ACM Transactions on Database Systems: The leadership of ACM SIGMOD and TODS have decided to offer the authors of certain TODS papers the opportunity to present their paper at the “next” SIGMOD conference. This agreement aims to make it more attractive to members of the SIGMOD community to publish in TODS, as well as to further enrich the technical program at the SIGMOD conferences.

Journal and conference publication differ in a number of respects. In the following, I review important differences, from the perspective of journal publication, and present a case for publication in TODS.

World One: Journal Publication. When a submission is received for consideration of publication in TODS, the submission is assigned to an Associate Editor who then is in charge of handling the submission and, in a sense, serves as the submissions ombudsman: The handling Associate Editor aims to do what is right for the submission and will take into account the authors responses to reviews. While the aim is to provide review results within 4 months, the journals review process can accommodate special circumstances as needed to get things right. For example, additional reviews can be obtained in a review round, and an additional round of reviewing can be introduced.

The traditional conference review process has a fixed schedule of deadlines and does not offer this flexibility. Some conferences have tried to achieve some of the flexibility by allowing one round of revision. Some conferences have also introduced procedures that may be viewed as a means of approximating the Associate Editor role as found at journals. They have introduced program committee vice-chairs and meta-reviewers, and they have introduced author feedback.

In my experience, these innovations to the conference review process are valuable but do not combine to yield the benefits of the journal review process. Specifically, what I call “hit-and-run” reviews still occur at times. These are superficial reviews that simply reject a paper without offering specifics. Key reasons why such reviews occur is that they are fast to do and that reviewers know that they can get away with them because there is no time for iteration. And I believe that the vice-chair and meta-reviewer roles are not always effective, a key reason also being tight deadlines. When serving in those roles, one often has to make accept/reject recommendations with the information already available.

As another difference between the journal and conference review processes, the Associate Editor who handles a submission recruits the reviewers from the global population of researchers, not from a fixed program committee that represents a fraction of the community. Admittedly, the program committees of top conferences consist of excellent researchers, and, even with the possibility of recruiting from the entire population, it is not always possible to recruit the top experts as reviewers of a particular journal submission.

The characteristics of the journal review process described here enable a more careful journal review process. The additional iterations that characterize the journal review process generally yield improvements to the results being published and their presentation.

Next, TODS papers are allowed to be 45 pages long in TODS format, and accompanying electronic-only appendices with no length restrictions are allowed. This arrangement affords authors exactly the space they need in order to present their results as best as possible. 

Further, accepted TODS papers are available online at the TODS website when they are ready; they are not hidden until they appear in a journal volume (or until a conference proceeding is made available, in the conference setting).

World Two: Conference Publication. However, conference publication also carries benefits inherently not available with traditional journal publication. When publishing in conference proceedings, the authors get to present their paper at a conference, typically either in a talk, as a poster, or both. And conference publication often means that the authors get funding that allows them to attend the conference, which then also allows them to attend other talks and to network with colleagues. Conference presentation also increases the visibility of the work.

TODS Paper Presentation at SIGMOD. The new agreement with ACM SIGMOD means that authors of TODS papers also get these benefits! The eligible papers include original research papers (i) that are not extensions of conference papers, (ii) that were accepted or published approximately within the year preceding the particular SIGMOD conference, and (iii) that were not already offered this opportunity at the previous conference. Focused surveys, critiques, corrections, and similar type papers are not eligible. These limitations mean that only entirely new, full papers are accommodated.

The presentation of a TODS paper at SIGMOD is in the form of a poster. As the data management community grows, more and more papers will be accepted at conferences such as SIGMOD. This means that there either has to be more parallel sessions with talks or that only some papers can be selected for presentation in the form of a talk. My guess is that poster presentation will gain in prominence in the years to come.

The agreement can be found in the “Policies” section of TODS’s website.

I hope that the opportunity to present TODS papers at SIGMOD will attract additional submissions to TODS.

Acknowledgments: Walid Aref and Divesh Srivastava provided helpful suggestions. However, they should not be held accountable for the views expressed here. ?

Christian S. Jensen

All ACM Journals | See Full Journal Index