The SPLC challenge track aims at providing the software product-line (SPL) community with a set of case studies that tackle relevant SPL problems and challenge the state of the art. The challenge track was established with SPLC 2018 in Gothenburg, initiated by Sarah Nadi, Timo Kehrer, and Thorsten Berger. The challenge is organized in two phases every year. The first phase is the call for cases. Submitted cases will be reviewed by the program committee to ensure that all required information is clearly described. Accepted cases will be part of the official conference proceedings. Authors of the accepted cases must attend the challenge track and participate in the discussion. The second phase is a call for solutions to both newly accepted and previously accepted cases. Submitted case solutions will be peer-reviewed by the challenge program committee. Accepted solutions will also be part of the official conference proceedings. Authors of accepted solutions must present their papers during the conference.

This year, we are placing a special emphasis in our call for cases and solutions on the need for actionable benchmarks and competition, akin to those prevalent in the machine learning and SAT communities, and as developed on platforms like Kaggle. We believe that introducing such benchmarks and competition will significantly enhance the ability to compare state-of-the-art solutions and new contributions in the field of SPL engineering. This kind of proposal should not only foster a spirit of innovation and collaboration but also provides a concrete, measurable framework for assessing the effectiveness and efficiency of new methodologies or implementations. We encourage submissions that offer well-defined, challenging, and replicable benchmarks, thereby setting the stage for a more lively and results-oriented challenge track.

Important Dates (AoE time)

• Challenge case submission: March 13, 2024
• Challenge case notification: April 4, 2024
• Camera-ready version of challenge case: April 11, 2024
• Call for challenge solutions released: April 12, 2024
• Challenge solution submission: June 10, 2024
• Challenge solution notification: July 1, 2024
• Camera-ready version of challenge solution: July 10, 2024

The idea of the challenge track is to provide participants with a set of case studies that tackle relevant SPL problems and challenge the state of the art. The challenge track happens in two phases.

• In the first phase, there will be a call for cases. The program committee will review the submitted cases to ensure that all required information is clearly described. Accepted cases will be part of the official conference proceedings. Authors of the accepted cases must attend the challenge track and participate in the discussion.
• In the second phase, there will be a call for solutions to the accepted cases and cases from previous years. Submitted case solutions will be peer-reviewed by the challenge program committee. Accepted solutions will also be part of the official conference proceedings. Authors of accepted solutions must present their papers during the conference.

You can check the challenges from previous years on the official website of the challenge track: https://variability-challenges.github.io/

Proposed challenges should be kept simple, as too complex ones may discourage people from trying to propose solutions. A challenge can be new, but it is also possible to update and improve existing challenges. For example, there could be new datasets or automated tools for benchmarking that deserve to expand or refine an existing challenge.

A case description can be a particular dataset accompanied by an answer to the following questions:

• What is your dataset, and how was it obtained?
• What is the size of the dataset?
• How can this data be accessed?
• Do you provide tools to process the data?

Furthermore, provide at least one concrete question you want participants to answer. You can provide multiple questions. For each question, provide the criteria for evaluating an answer/solution.

A case description can also be a call for a solution to a specific problem in a given context, which should answer the following questions:

• What is the concrete problem you want participants to solve?

• How can a solution be evaluated? The following are some suggestions on how to specify evaluation criteria

• Concrete evaluation metrics (e.g., precision, recall, accuracy etc. depending on the problem)

• Concrete test cases participants can evaluate their solution with (e.g., provide inputs and expected outputs and participants are expected to provide a solution that gets from one point to the other)

• A list of existing solutions to evaluate their solution against

• A reference implementation to compare against, according to particular metrics

In both case descriptions (either a problem or a solution), there are some requirements to fulfill:

• The description should contain the URL of a public repository or artifact page containing all the instructions needed to initiate the case study.

• The description should describe the requirements for the solution, or a requirement specification document should be published in the repository. The requirements’ description should help researchers who want to build the solution of or answer to the specific questions.

• Optional: Case authors may include a list of 5 senior PhD students or junior researchers (postdocs or junior professors) who have the expertise required to evaluate submitted solutions. The case authors themselves may be part of this list. The challenge track co-chairs will consider this list when creating the SPLC challenge program committee.

After the description of accepted cases is finalized, a call for solutions will be released. Please note that case authors can submit a solution to their own challenge.

List of Topics

systems of interest: software product lines, configurable systems, evolving systems, adaptive systems, problem-space variability, solution-space variability

techniques of interest: machine learning, automated reasoning, constraint programming, reverse engineering, performance prediction, testing, fault localization, model checking, model-driven engineering

The listed topics are not exclusive. If you are unsure whether your challenge is in scope, please get in contact with the track chairs: splc2024challengetrack@easychair.org

Committees

Track Chairs

Mathieu Acher

University of Rennes, France

Thomas Thüm

University of Ülm, Germany

Program Committee

• Mohammad Reza Mousavi, UK
• Son Nguyen, Vietnam
• Wesley K. G. Assunção, USA
• Xhevahire Ternava, France
• Eduardo Figueiredo, Brazil
• Jacob Krüger, The Netherlands
• Javad Ghofrani, Germany
• Michael Lienhardt, France
• Shiyi Wei, USA

Publication

Accepted cases as well as accepted solutions will be part of the official proceedings of SPLC 2024

All questions about submissions should be emailed to splc2024challengetrack@easychair.org