View on GitHub

Causal Methods in Software Engineering (CauSE 2027)

Workshop CauSE 2027

Description:

This second edition of the CauSE workshop is organized in the context of ICSE 2027, Dublin, Ireland (25 April – 1 May 2027).

Motivation:

Despite their potential, causal methods have not yet been broadly leveraged by the software engineering community. While preliminary studies demonstrated their benefit in specific areas, their broad and systematic exploitation for software engineering is still far from coming. The objective of this workshop is to provide a platform for participants to share their research, experiences, and insights on causal inference methods and their applications in software engineering. The broader aim is to foster networking and open new collaboration opportunities, encouraging the development of a strong community.

This workshop aims to bring together researchers, practitioners, and students to explore the growing field of causal inference and, more broadly, causal AI (including causal discovery, mediation analysis, counterfactual analysis, root-cause and causal attribution analysis) in software engineering. Causal inference methods allow the identification and estimation of causal effects from observational data, distinguishing between spurious correlations and causal effects. Despite a growing interest in the topic, the application of such methods in software engineering is not widespread, and there is a need to foster a community to exchange and promote scientific activities on the topic. The spread of LLM-based and agentic systems makes this need more pressing, as such systems are largely assessed through correlational benchmarks while the questions they raise — what caused a failure, what would have happened under a different design choice — are causal ones.

Workshop goals:

The workshop will provide a platform for participants to share their research, experiences, and insights on causal inference methods and their applications in software engineering. It will also facilitate networking and collaboration opportunities, encouraging the development of a strong community.

Call for papers:

Topics of interest:

The workshop intends to keep the scope of the application use cases as broad as possible. We don’t want to restrict the type of causal methods applied and want this workshop to be an open stage for SE researchers to discuss which causal approach fits best a given use case. The types of work expected include (but are not limited to) proof of concept, benchmarks, empirical studies, lessons learned reports, literature reviews, etc.

Topics include the application of causal reasoning methods, such as causal discovery, causal inference, and the causal treatment of machine learning (Causal Machine Learning, Causal Reinforcement Learning), as well as the use of large language models as instruments for causal modelling, to:

Manuscript information:

Submitted papers should present original, unpublished work, relevant to one of the topics above. CauSE 2027 will accept:

Submissions must be in English and in PDF format. At the time of submission, all papers must conform to the ICSE 2027 format and submission guidelines. The workshop will employ a double-anonymous review process. All submissions will be refereed by at least three members of the program committee. Accepted submissions will be published in the ICSE 2027 companion proceedings. At least one author of each accepted paper is required to register for and present at the workshop.

Important Dates:

Workshop Program:

Program Committee:

Organizers and contacts

Neil Walkinshaw (contact page, google scholar) is a Senior Lecturer at the University of Sheffield. His research focuses on software quality assurance, particularly “black-box” components, and he specializes in applying Machine Learning and data analysis algorithms to testing, reverse-engineering, and safety-case assessment. He received a grant from CITCoM (2021-2024) for the project “Causal Inference for Testing of Computational Models.”

Luca Giamattei (contact page, google scholar) is researcher at the Federico II University of Naples, working in the Dependable Systems and Software Engineering Research Team (DESSERT). His research interests encompass the use of causal reasoning in software testing.

Julian Frattini (contact page, google scholar) is a postdoctoral researcher at the Chalmers University of Technology, Sweden. His research interests revolve around quality of requirements artifacts, empirical research methods with a focus on statistical causal inference and Bayesian data analysis.

Hans-Martin Heyn (contact page, google scholar) is a Senior Lecturer at the Computer Science and Engineering Department of the University of Gothenburg and Chalmers University of Technology in Sweden. His research focuses on Software Engineering for AI and distributed cyber-physical systems. He is especially interested in studying how to derive requirements for training data and data at runtime based on prior knowledge about the problem domain. A research direction that triggered his interest for achieving this aim is causal modelling and information entropy.