ACM SIGMOD/PODS Conference: Vancouver, 2008
SIGMOD: Call for Papers

The annual ACM SIGMOD conference is a leading international forum for database researchers, practitioners, developers, and users to exchange ideas and results. We invite submissions of original research contributions, case studies, and industrial papers, as well as proposals for demonstrations and tutorials.  We encourage submissions relating to all aspects of data management defined broadly, and particularly encourage work on topics of emerging interest in the research and development communities.

Areas of interest include but are not limited to the following:

• Benchmarking and performance evaluation
• Case studies that reveal research challenges
• Data quality, semantics and integration
• Database monitoring and tuning
• Data privacy and security
• Data mining and machine learning
• Database tuning
• Decision support on large data sets
• Embedded, sensor and mobile databases
• Indexing
• Managing uncertain and imprecise information
• Novel/Advanced applications, systems, and platforms
• Peer-to-peer and networked data management
• Personalized information systems
• Query processing and optimization
• Replication, caching, and publish-subscribe systems
• Semi-structured data
• Storage and transaction management
• Text and image databases
• Large scale social networks
• Web services

Because SIGMOD is a very selective conference, we would like to make explicit some of our philosophical principles.

1. SIGMOD is a great conference, but it's still a conference. This means that we are interested in innovative ideas that are well presented but are not necessarily perfect. If a definition is poorly written but can be made right fairly easily or if an experiment is missing, but could be done in short order, but the paper is otherwise of great interest, then we will work with authors to correct the problem. In other words, an innovative paper with some blemishes may still be accepted, even if less enthusiastically than otherwise.
2. Conversely, tiny, one-could-have-thought-of-that-in-an-afternoon variants on existing algorithms or languages are not interesting unless they produce an order of magnitude or large constant factor performance or other quality improvement, preferably on a standard body of data and with the code available to ensure reproducibility.
3. Similarly, if a technique is proposed to solve a problem and ignores (or superficially dismisses) an entire subfield of computer science or statistics that deals with that problem, then that might be a reason to reject. We expect the main previous references to be treated carefully so that the paper's contribution can be made clear.
4. Case studies that could inspire new avenues of research are of definite interest. Case studies that merely recount the deployment of a database management system in a straightforward way are not.
5. Even the best ideas can be misunderstood if the writing is poor, so we encourage all authors to ensure that their submissions are written in clear and grammatically correct English.
6. For papers that perform quantitative performance experiments, the experiments should describe a variety of scenarios in order to give a full picture of the authors' techniques' impact. Thus:

• The experimental results should include cases when the proposed techniques help as well as those where they don't or they are detrimental (if any). Authors are best suited to identify and test these scenarios in such a way that reviewers will not feel that some important situations have been left out. For example, a proposal for a new index structure should test update as well as query scenarios.
• Authors should ensure that the numbers produced are statistically meaningful (e.g., the number of runs and the standard error should be specified).
• Whenever possible, standard datasets should be used, either standard benchmarks or the datasets used in previous papers to which the authors compare their work. These may be supplemented by the authors' own datasets.
• Descriptions of experimental evaluations should describe all parameters that impacted a measure's outcome: hardware and software configuration, data sets, the precise operations being measured etc.

Manuscript preparation and submission guidelines

Detailed guidelines for research papers, formatting instructions, and submission instructions can be accessed through the links on the right panel of this page. No late submissions or and no submissions via email (or hardcopy) will be accepted. For other form of submissions (demo, industrial track talks, tutorials, panels) please see the relevant instructions from the conference home page.

Rollover Paper Policy

If your paper was invited to be a rollover paper from VLDB 2007 to sigmod 2008, please do the following: (i) be sure to follow the normal SIGMOD anonymization rules for the paper (ii) use three to five additional pages as an appendix to explain point by point how you dealt with the VLDB reviews. We want you to respond to every suggestion made by the reviewers, but you should make no other technical points.

Send this appendix separately to shasha@cs.nyu.edu. by 3 PM US Eastern Time November 16, 2007. This should NOT be anonymous. The subject line of your email should identify the SIGMOD paper number. This appendix should NOT be part of your submission through the CMT site.

Questions about any of the requirements or reviewing process listed on this page should be directed to the SIGMOD 2008 Program Committee Chair Dennis Shasha (shasha@cs.nyu.edu).