Database Publication Practices: Recap of SIGMOD 2004 Panel

1
Database Publication PracticesRecap of SIGMOD
2004 Panel

• Zachary G. Ives
• University of Pennsylvania

October 22, 2015 VLDB 2005, Trondheim, Norway
2
Observed Problems

• Conferences are being flooded
• Lack of dis-incentives against not-ready-for-prime
  -time submissions
• Common practice to re-submit every rejection to
  the next conference not always with major fixes
• Requires immense program committees
• Result has been a lack of quality control or
  consistency in reviews
• Not only frustrating, but can harm careers!

3
Possible Solutions

• Two classes
• Make it harder to submit (or re-submit) papers
• limit number of submissions break SIGMOD gt
  VLDB gt ICDE pipeline
• Or streamline and improve the review process
• (see SIGMOD04 proceedings SIGMOD Record
  summary, and VLDB05 conference proceedings for
  more details than I have time for here)

4
Ideas for Improving the Review Process

• Increase quality control and participation in
  reviews
• Open reviewing on the Web
• More reviewers per paper
• Two-tier PCs (as with ICDE)
• Mandatory physical PC meetings
• Reviewer evaluations
• Hybridize journals (more effective reviews) and
  conferences (prestige visibility)
• Institutional memory (opt-in) for borderline
  rejections in conferences similar to journals
  consistent reviewers rebuttal
• Accelerated journal reviews, mechanisms for
  prestige
• Merging of journals and conferences use the
  conference as a forum for presenting the most
  interesting papers from the past year