Should Conferences Meet Journals and Where?: A Proposal For 'PACM'

For many months now, the ACM Publications Board Conferences Committee has been working on a proposal that brings together conference and journal publishing. Here, we offer some background leading up to this proposal, summarize the plan, and solicit your input. We are also pleased to present two leading computer scientists—Kathryn S. McKinley and David S. Rosenblum—who argue for and against the proposal beginning on page 43.

This proposal has been driven by strong input from the CS research community where the prevailing feeling is, despite strong review processes and selective acceptance, publishing in conference proceedings puts CS researchers at a disadvantage with respect to researchers in other scientific disciplines, where journal publication predominates, and is hence fundamentally detrimental to the CS field.

While some claim U.S.-based efforts by the National Research Council and the Computing Research Association have helped make top-quality conference papers "count" as much or more than peer-reviewed journal papers, others claim conference papers still have second-class status. Indeed, funding and merit systems in many countries do not fully recognize conference papers. Many believe the conference-centric culture of CS puts researchers at a disadvantage when competing with researchers in other science disciplines for top science awards and career progression.

The relationship between conferences and journals is complex. Many computer scientists lament the inherent flaws and limitations of conference publishing: hard deadlines, page limits, limited review and revision time, and overloaded review committees. Attending conferences comes at a cost, limiting access. And some of the initial reasons CS publishing favored conferences—the lack of CS journals, the slow speed of journal publishing, and the high cost of journal papers—have mostly disappeared. Still, many in our community argue that conferences are where the most exciting research is published; they seek to maintain the vibrant exchange that happens there. For this reason, we charged a diverse task force to explore whether the field would be better served through new models of conference-journal interaction.

Back to Top

What Is the Status Quo?

ACM currently offers three ways to publish conference papers in journals.

Revision of the paper by adding at least 25%–33% new content and submitting to a journal for review. Disadvantages include "citation splitting" and lag time.

Journal-first publication where the paper is reviewed by the journal and published there, but the authors are invited to present at the conference. This option is the basis of the successful HiPEAC conference, whose papers are submitted directly to ACM TACO, and is increasingly common as an option in several SIG-sponsored conferences.

Journal-integrated publication where the conference integrates its committee review process with a journal. Papers accepted by the conference-review process are published in the journal and presented at the conference. Papers that need additional rounds of review are transferred to the journal-review process. Papers accepted through the journal-review process can be presented at a later conference. This model is used by SIGGRAPH in conjunction with ACM TOG.

Back to Top

New Proposal: A Proceedings-Focused Journal Series

The committee proposes a fourth alternative—a journal series specifically created to publish the proceedings of ACM’s highest quality conferences. Tentatively called Proceedings of the ACM, it would parallel ACM Transactions with a set of journals publishing high-quality research vetted by research communities through conferences. A few key features and criteria:

• A series would cover several conferences related by SIG or theme (for example, PACM Programming Languages or PACM Distributed Systems).
• All included proceedings papers would be reviewed for correctness, accuracy, completeness, and impact through a selective peer-review process. This process would require written reviews by a minimum number of three qualified reviewers. Documented policies would ensure the integrity of the review process.
• The review process must permit at least one substantive revision by the authors and review of that revision by the reviewers to determine whether their concerns are adequately met.
• Papers may not have artificial limits on length that would prevent full disclosure of references, documentation of methods, and so on. Supplemental materials would be clearly labeled as to whether they underwent peer review.
• PACM series may invite submissions or second and subsequent revisions outside the conference timeline and review cycle.
• Inclusion of a conference proceeding in a PACM series is subject to initial and periodic review by the series steering committee and the ACM Publications Board.

We want to hear from you! What do you think? Please read the arguments for and against this proposal on p. 43 and tell us what you think at https://www.surveymonkey.com/r/PACM2015 by Sept. 20 if possible!

Back to Top