Workshop Overview
The Archive '10 workshop will focus on the creation of archives of computer-based experiments: capturing and publishing entire experiments that are fully encapsulated, ready for immediate replay, and open to inspection. It will bring together a few areas of the scientific community that represent fairly advanced infrastructure for archiving experiments and data (physicists and biomedical researchers) with two areas of the computer systems community for which significant progress is still needed (networks and compilers). The workshop will also include experts in enabling technologies and publishing.
- Dates: May 25-26, 2010
- Venue: Warnock Engineering Building, University of Utah campus, Salt Lake City, UT, USA. See the TravelInformation.
- Program: See the WorkshopSchedule.
Participation is by invitation. For more information, contact the WorkshopOrganizers.
Motivation
Significant advances in computing infrastructure have come from experimental research that measures performance, throughput, accuracy, energy requirements, and other desirable features, and attempts to compare new approaches to the state of the art. Unfortunately, often it is infeasible for researchers other than the original developers to repeat these experiments, or provide valid comparisons with prior work, which has a profoundly negative impact on moving research ideas into practice. In the current era of large data centers, advances in data-mining technology, and increasingly powerful computers, we believe that technology advances could be applied to archiving experimental data to dramatically advance scientific research as well as the computing field.
This workshop will focus attention on this issue to take the next step in publishing and archiving computing and other scientific research: to capture and publish entire experiments that are fully encapsulated, ready for immediate replay, and open to inspection. Experiments and results would be available to be examined, repeated, extended, and reused by anyone. Experiment repositories will connect published research to its sources in ways that support new analyses and new executions of computer-based systems. These archives would directly benefit many research and education communities, and their construction would involve a wide array of systems research challenges.
This workshop will ascertain the breadth and depth of various communities' interest in open community archives of replayable experiments. A long-term goal is to build a lasting coalition of stakeholders that includes computer and computational scientists, educators, librarians, publishers, and professional societies that will contribute to the eventual construction and operation of an open community archives of replayable experiments. We observe that certain scientific communities are more advanced in their archiving of experiments and resulting data than the computer science systems research community. In some ways, this reality is surprising given that computer science systems research has enabled these scientific communities to develop their experimental archives!
In this workshop, we plan to bring together a few areas of the scientific community that represent fairly advanced infrastructure for archiving of experiments and data (physicists and biomedical cancer researchers) with two areas of the computer science systems community for which significant progress is still needed (computer networks and compilers). A 2007 NSF-sponsored compiler research workshop found the lack of experimental repeatability was hampering the field. (This is described in a February 2009 Communications of the ACM article.) Because the systems community has some additional technical challenges to solve, we will also invite experts in enabling technologies and publishing.
The desired product of this workshop will be a vision of required infrastructure and cultural changes in the focus research communities for archiving and publishing experimental results.
Support
The Archive '10 workshop is supported by the National Science Foundation.
This material is based upon work supported by the National Science Foundation under Grant No. 0709430. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.
