ERROR 2023, Limassol, Cyprus

3rd Workshop on E-science ReseaRch leading tO negative Results
Tuesday, Oct 10, 2023 at 14:00, MEGARON A

ERROR 2023 provides the e-Science community a dedicated and active forum for exchanging cross-discipline experiences on research leading to negative results and lessons learned. The workshop covers both applications and systems areas, including topics in research methodology, reproducibility, the applications/systems interface, social problems in computational science, and other relevant areas. Zoom link: https://www.zoomgov.com/j/1604968445?pwd=c1NaT1o5S01yK0xEMnlwbmNQMDl2Zz09

Info about past ERROR workshop may be found here.

In conjunction with  
Proceedings with

Workshop Program

Oct 10, 14:00, MEGARON A

Time Event
2:00-2:05pm Welcome & Introduction
Ketan Maheshwari
2:05-2:30pm Negative Results in Cloud, Edge and HPC Science
Beth Plale, Indiana University Bloomington
2:30pm-2:40pm Scientific data reduction: benefits and lessons learned
Qian Gong, Oak Ridge National Laboratory
2:40pm-2:50pm Lessons learned from developing Automated Machine Learning on HPC
Romaine Egele, Université Paris-Saclay
2:50pm-3:00pm Human error: reference out of scope
Adam Craig, Hong Kong Baptist University
3:00pm-3:10pm Opportunities and Challenges in Real-world System Deployments
Keshav Kaushik, Birla Institute Of Technology And Science at Pilani
3:10pm-3:20pm When the not-so-smart manipulation of smart pointers kills performance
Fred Suter, Oak Ridge National Laboratory
3:20pm-3:30pm Benchmarking the Bruck Algorithm for All-to-All Communication in multi-GPU environment using NCCL
Andres Sewell, Utah State University
3:30pm-4:00pm Break
4:00pm-5:00pm Discussion
Lead By Daniel S. Katz

Important Dates

  • July 7, 2023

    July 14, 2023

    Papers Submission
  • August 8, 2023

    Paper Acceptance Notifications
  • August 14, 2023

    Camera-ready Submissions
  • October 10, 2023

    Workshop

All deadlines are Anywhere on Earth (AoE).

Call for Papers

Escience researchers often invest significant time and effort in projects that can have unexpected outcomes. When results are obtained that do not fit the established molds expected by publishing venues, these results can be lost. The ERROR Workshop invites research results from computing and computational projects that encounter unexpected or negative results that would be difficult to report on elsewhere.

Novel hardware technologies and machine learning approaches (among others) are rapidly changing approaches, methods, and scale applied in escience domains. Researchers must deal with this novelty in multiple dimensions, many of which are beyond their control. Consequently, it is likely that some of the obtained results will not be of the expected form: they are negative (deviating from initial hypothesis), abnormal (anomalous to results from similar studies), or otherwise not useful in the established sense.

Under normal circumstances, such negative results and why they were obtained are seldom discussed, analyzed and published. Useful lessons are thus lost to the scientific community. Yet ignoring such results and the process by which they were obtained poses a risk of repetition. The fact that other researchers likely face the same situations and the same pitfalls further increases the cost of research, a cost that would have been avoided if the negative results were brought forward and discussed in-depth within and across communities.

Topics for the workshop

  • Unforeseen technology / problem / technique misfits
  • Institutional policies (on rejected research)
  • Failures and obstacles faced during a successful research work
  • Controversial results because of undiscovered technical glitch
  • Unconventional results which contradict theoretical expectations
  • Discovery of better approaches after significant efforts spent on research
  • Inadequate or misconfigured infrastructure
  • Abnormal and anomalous results
  • Ongoing research with setbacks and lessons learned
  • A hypothesis with one or more limiting assumptions found to be incorrect
  • Discovery of unexpected behavior in hardware, networks or platforms
  • Data size that is too big or too small for the applied technique
  • Implementation of simulation tools based on incorrect physical observations
  • Defect in software design, architecture and/or user interface
  • Software and platform incompatibilities
  • Zero defect software policy and its implications
  • Other topics relevant to the description above

Proceedings Publication

Accepted papers from the workshop will be published as part of eScience 2023 proceedings to be published by the IEEE Computer Society Press, USA and made available online through the IEEE Digital Library.

Paper Submission Guidelines

Authors are invited to submit a maximum of 6-page manuscripts describing original and unpublished work surrounding the aforementioned topics. The format of the paper should be of double column text using single spaced 10 point size on 8.5 x 11 inch pages, as per IEEE 8.5 x 11 manuscript guidelines. Templates are available from this link.
Authors should submit a PDF file that will print on a postscript printer to the easychair conference system.

Organization

Ketan Maheshwari (km0@ornl.gov)

Oak Ridge National Laboratory, USA

Justin Wozniak

Argonne National Laboratory, USA

Daniel S. Katz

University of Illinois at Urbana-Champaign, USA

Program Committee

Tainã Coleman

University of Southern California, USA

Arnaud Legrand

CNRS, France

Ulf Leser

Humboldt-Universitaet zu Berlin, Germany

Loïc Pottier

University of Southern California, USA

Tanu Malik

DePaul University, USA

Raül Sirvent

Barcelona Supercomputing Center, Spain

Sean R. Wilkinson

Oak Ridge National Laboratory, USA

Kathryn Knight

Oak Ridge National Laboratory, USA

Ankit R. Patel

University of Minho, Portugal

Olga Kuchar

Oak Ridge National Laboratory, USA