About the Workshop
AI-enabled software and AI-assisted engineering require explanations that support justified decisions. Yet a faithful explanation may not enable stakeholder understanding, and improved understanding may not improve an engineering task. These relationships must be specified and validated when explainability is treated as a first-class software quality attribute.
EXQA treats this as an engineering problem organized around three distinct but interacting concerns:
Faithfulness
Does the explanation accurately reflect the system's actual behavior or decision process?
Understanding
Does the explanation enable a stakeholder to construct an accurate mental model, draw correct task-relevant inferences, and recognize limitations?
Engineering usefulness
Does the explanation improve a concrete software engineering task or decision, such as model selection, defect diagnosis, patch review, or deployment, given its effort and cost?
The scope covers both explanations of AI-enabled system behavior and explanations of AI recommendations in AI-assisted engineering, including model selection, code review, testing, debugging, and architectural evolution. We welcome researchers and practitioners in requirements, architecture, testing, empirical software engineering, human factors, and AI engineering. Preliminary studies, industrial experience, replications, and negative findings are explicitly welcome.
Topics of Interest
Contributions may address faithfulness, stakeholder understanding, engineering usefulness, or their relationships within a clearly identified software engineering context. Topics include, but are not limited to:
- Explainability requirements and quality models: stakeholder needs, explanation requirements, quality-attribute scenarios, acceptance criteria, and relationships with other software quality attributes.
- Architectures for explainable software systems: architectural patterns, explanation services and interfaces, evidence provenance, compositional explanations, and system-level explainability across models, services, and agents.
- Explanation methods, representations, and tools: explanation generation, abstraction, summarization, visualization, interactive explanations, and integration of explanatory support into development and operational environments.
- Measurement, verification, and testing: faithfulness, validity of explanation metrics, uncertainty, robustness, consistency, explanation testing, and evaluation protocols connecting technical measures to stakeholder and engineering outcomes.
- Stakeholder understanding and human interaction: mental models, expertise, cognitive load, communication of assumptions and limitations, stakeholder-specific explanations, and human–AI collaboration.
- Explainability for AI-assisted engineering tasks: requirements analysis, code generation and review, test generation, debugging, vulnerability analysis, architecture decisions, maintenance, and explanations of agent actions and recommendations.
- Engineering decisions and quality trade-offs: explanation-supported model selection, configuration, integration, and deployment; multi-objective decision support; and trade-offs involving accuracy, reliability, latency, cost, privacy, and security.
- Explainability throughout software evolution and operation: monitoring, explanation drift, regression testing, versioning, evidence retention, revalidation, runtime adaptation, and governance of explanatory artifacts and responsibilities.
- Empirical studies and shared research resources: industrial experience, developer studies, longitudinal evaluations, benchmarks, datasets, reusable engineering scenarios, replication, and negative results.
Contributions need not address all three concerns. Each submission should identify the explanatory object, the engineering task, the evidence offered, and any untested assumptions.
Submission Categories
All page limits include references.
| Category | Length | Description |
|---|---|---|
| Full research papers | Up to 8 pages | Substantive conceptual, methodological, technical, or empirical contributions that advance the workshop's research agenda. |
| Short research and new ideas papers | Up to 4 pages (extended abstract) | Preliminary findings, emerging approaches, position statements, and research visions that articulate a clear problem and provide a productive basis for discussion. |
| Tool demonstration papers | Up to 4 pages (extended abstract) | Tools or prototypes supporting the specification, implementation, evaluation, or maintenance of explainability. Describe the intended users, engineering workflow, distinctive capabilities, and planned demonstration. |
| Industrial and experience reports | Up to 8 pages | Practical adoption, deployment, or evaluation, including contextual constraints, observed outcomes, challenges, and lessons for researchers and practitioners. |
How to submit
- Submissions must use the ACM Primary Article Template (
sigconfformat), consistent with FSE 2027 requirements. - EXQA uses a double-anonymous review process. Remove author names and affiliations, and refer to your own prior work in the third person.
- Submissions of five pages or fewer (including references) are published as extended abstracts and are not subject to an article processing charge (APC).
- Submission site: TBA
Review and selection
Each submission will receive at least three reviews from conflict-free program committee members, followed by committee discussion. Selection considers relevance, clarity, originality, soundness, and potential to stimulate discussion, with expectations calibrated to the paper type.
Conflicts of interest
Organizers and their directly supervised students, postdocs, and employees may not submit to EXQA. The workshop follows the ACM SIGSOFT Policy on Program Committees. Conflicted reviewers do not review or take part in decisions on the affected paper; an unconflicted organizer or designated handling chair manages such submissions. Non-organizer PC members may submit under the same standards, with independent handling.
Important Dates
| Paper submission deadline | 11 February 2027 |
| Acceptance notification | 18 March 2027 |
| Camera-ready deadline (hard) | 8 April 2027 |
| Workshop | TBA (FSE 2027, July 12–16) |
All deadlines are 23:59:59 Anywhere on Earth (UTC−12).
Publication and Registration
Accepted papers will appear in the workshop proceedings in the ACM Digital Library, subject to the FSE 2027 publication process. At least one author of each accepted contribution must register for the workshop for the paper to be included in the proceedings.
Workshop Format
EXQA is a half-day, in-person workshop. At least half of the programmed time is devoted to discussion and collaborative work. The program includes:
- Opening, invited talks, and paper presentations. Two invited talks, one from academia and one from industry, followed by thematic paper sessions. Presentations should state an engineering problem, the contribution, and an open question for discussion.
- Thematic working sessions. Small groups organized around requirements, architecture, measurement, human interaction, or lifecycle management each develop a research question, engineering scenario, or evaluation approach.
- Panel and collective synthesis. Groups compare findings, discuss competing perspectives, and identify research priorities.
Authors of accepted papers will be invited to provide a short discussion question or scenario before the workshop. Intended outputs are a set of research questions, reusable engineering scenarios, and methodological recommendations.
Attendance is open to all registered FSE participants, including those without an accepted paper. Students and early-career researchers are especially encouraged to join.
Organizers
Yan Liu
Concordia University, Montreal, Canada
yan.liu@concordia.ca · Profile
Yan Liu is a Professor at Concordia University whose research spans cloud computing, distributed systems, software architecture, and explainable AI. Her explainability work connects software services, quality measurement, and architectural decisions. Relevant community activities include ICSE NIER, IEEE Software special issue guest editor, IEEE Transactions on Service Computing associate editor, CASCON observability and explainability workshops and the ICSA 2026 working session.
Zheng Li
Queen's University Belfast, Belfast, United Kingdom
zheng.li@qub.ac.uk · Profile
Zheng Li received his Ph.D. from the Australian National University (ANU) and his M.Phil. from the University of New South Wales (UNSW). During his doctoral studies, he was also a graduate researcher with the Software Systems Research Group (SSRG) at National ICT Australia (NICTA). He is currently a Senior Lecturer (Associate Professor) in the School of Electronics, Electrical Engineering and Computer Science at Queen’s University Belfast, UK. Previously, he was a tenured Assistant Professor in the Department of Computer Science at the University of Concepción, Chile, and a postdoctoral researcher with the Cloud Control group at Lund University, Sweden. He has also been a visiting research fellow at the Software Institute of Nanjing University, China. His research interests include big data analytics, edge and cloud computing, edge intelligence, empirical software engineering, and performance engineering.
Abdelwahab Hamou-Lhadj
Concordia University, Montreal, Canada
wahab.hamou-lhadj@concordia.ca · Profile
Abdelwahab Hamou-Lhadj is a Professor and Chair of the Department of Electrical and Computer Engineering at Concordia University. He was an Affiliate Researcher at NASA's Jet Propulsion Laboratory (JPL), Caltech, from 2022 to 2025. His research focuses on AIOps, software observability, the trustworthiness of AI systems, software engineering, and model-driven engineering. He has led numerous collaborative research projects with industry and institutional partners, translating advances in AI and software engineering into practical solutions for complex software ecosystems. Dr. Hamou-Lhadj has served on the program committees of many leading international conferences and has organized numerous workshops and scientific events. He is currently an Associate Editor of IEEE Transactions on Reliability and IEEE Software.
Qing Tan
Athabasca University, Athabasca, Canada
qingt@athabascau.ca · Profile
Qing Tan is a Full Professor in the School of Computing and Information Systems at Athabasca University. He received his PhD in Cybernetics Engineering for Robotics from the Norwegian Institute of Technology (now the Norwegian University of Science and Technology) in 1993. He subsequently served as a Foreign Senior Research Fellow at the Japan Atomic Energy Research Institute and completed a postdoctoral fellowship at the University of Alberta before joining Athabasca University in 2007. Dr. Tan’s research spans mobile learning, telepresence robotics, and cyber-physical systems. His current work focuses on the Telepresence Robot Empowered Smart Lab (TRESL) and the integration of Autonomous Digital Twin Teammates (ADTTs) into Human–Autonomy Teaming (HAT), exploring how artificial intelligence and digital twins can support transparent, adaptive, and trustworthy collaboration between humans and autonomous systems. He has extensive experience in teaching, student supervision, and research, with a strong record of scholarly publications and externally funded projects.
Zhen Yang
Shandong University, Qingdao, China
zhenyang@sdu.edu.cn · Profile
Zhen Yang is an Assistant Professor in the School of Computer Science and Technology at Shandong University. His main research areas are software engineering, program comprehension, and program generation. He received his PhD in Computer Science from City University of Hong Kong in October 2023, supervised by Prof. Jacky Keung. He has published nearly 40 papers in leading international software engineering journals and conferences, including FSE, ASE, ISSTA, TSE, and TOSEM, with over 1,400 Google Scholar citations. He has led projects funded by the National Natural Science Foundation of China, the Shandong Provincial Natural Science Foundation, and others. He was selected for the Shandong Provincial Young Science and Technology Talent Support Program and the Shandong Provincial Overseas Science and Technology Talent Introduction Program. He serves as a program committee member for conferences such as SANER and MSR, and as a reviewer for software engineering journals such as TOSEM and TSE.
Jun Huang
Concordia University, Montreal, Canada
jun.huang@mail.concordia.ca · Profile
Jun Huang is a Ph.D. student in Electrical and Computer Engineering at Concordia University, Montréal, Canada. Jun’s research is on explainable AI (XAI): the evaluation of XAI methods, provenance-based machine learning and explainability operations (MLXOps), and explainability for medical imaging and software vulnerability analysis. Jun has published at ICSE-NIER, IEEE Big Data, and COMPSAC, and in IEEE Transactions on Cloud Computing and IEEE Access. Jun presented the talk “How do we efficiently integrate explainable AI (XAI) to help the deep learning of medical images?” at the 1st International Workshop on Observability and Explainability of AI Systems (OXAI’24), co-located with CASCON 2024 in Toronto.
Program Committee
To be announced.
Contact
Questions about the workshop: yan.liu@concordia.ca