About ISE'26

The Fifth Brazilian Workshop on Intelligent Software Engineering (ISE'26) will be held in person. ISE'26 is co-located with the 17th Brazilian Conference on Software (CBSoft 2026).

Aims and Scope

In modern society, software is ubiquitous, being present in almost every aspect of life. However, developing software is costly. Therefore, there is a continuous effort for innovative designs for helping make software more reliable, maintainable, and reduce its development cost. Further, there is an increase in AI-enabled software, bringing new software engineering challenges.

In recent years, Intelligent Software Engineering (ISE) has emerged as a promising means to address these challenges. ISE is ambidextrous, including solutions based on (i) applying Intelligent Techniques to Software Engineering problems and (ii) applying Software Engineering to developing Intelligent Systems.

An Intelligent Technique is defined as a technique that explores data (from digital artifacts or domain experts) for knowledge discovery, reasoning, learning, planning, natural language processing, perception, or supporting decision-making. Examples of Intelligent Techniques are search and optimization (e.g., Genetic Algorithm), Machine Learning, Deep Learning, Data Mining, Recommender Systems, Reasoning Under Uncertainty (e.g., Bayesian Networks), Software Analytics, and Decision Analysis. An intelligent system is a system that applies an Intelligent Technique for a given domain, such as a bug prediction, code recommendation, code quality analysis, code refactoring, software testing automation, team formation, requirements analysis, software project management, risk management and continuous integration and deployment.

The goal of ISE'26 is to strengthen the Intelligent Software Engineering community by integrating researchers and professionals from different areas (Software Engineering, Artificial Intelligence, Analytics, etc) to discuss and advance the state of the art and practice of ISE, its use, and application in the industry.

Topics that are within the scope of this workshop include (but are not limited to):

Submission Guidelines

Paper Format

ISE accepts papers in two formats.

Technical Papers (6 pages)

- ISE accepts papers within the context of "Intelligent Software Engineering". Positive and negative results are welcome, though negative results should still be based on rigorous research and provide details on lessons learned.

Industry papers / Emerging Results (2-4 pages)

- Results, challenges, lessons learned from industrial applications of "Intelligent Software Engineering".
- New and inspiring ideas, as well as ongoing research with preliminary and interesting results on applications of "Intelligent Software Engineering".

Instructions for Authors

Papers can be submitted in Portuguese or English. Submission in English is strongly encouraged. All submissions must be in Adobe Portable Document Format (PDF) and must comply with the ACM 2-column SBES format available at this overleaf project link or latex files (.zip) download. LaTeX users must use the acmart.cls class provided in the template with the conference format enabled at the document preamble: \documentclass[sigconf]{acmart}

Authors must use the ACM-Reference-Format.bst bibliography style also provided in the template: \bibliographystyle{ACM-Reference-Format}

Submissions must be no longer than the limits defined in the Paper Format section, including all figures and references.

Papers should be electronically submitted through JEMS. Submissions that do not comply with the required submission format, are out of the scope of the symposium or were submitted or published in any other forum (conference or journal) will be rejected without review. At least two Program Committee members will review all other papers.

Open Science Policy

ISE'26 supports the principles of Open Science, aiming to increase transparency, reproducibility, and collaboration in research. Authors are encouraged to share data, code, and other research artifacts in an anonymized and curated format to help reviewers and readers validate results and reuse materials. We understand that in some contexts—such as qualitative or industrial studies—sharing data may not be possible due to privacy, confidentiality, or other restrictions.

Policy on the Use of Generative AI Tools

Authors submitting to ISE'26 must follow ethical guidelines regarding the use of AI tools in research writing. It is not allowed to:

It is allowed, with proper acknowledgment in the paper, to:

Example: "ChatGPT was used to revise Section 2 and generate the initial version of Table 1."

It is also acceptable (without the need to mention it) to:

At the time of submission, authors will be required to disclose whether AI tools were used during the preparation of the manuscript. This information is requested strictly for ethical transparency, and authors should not feel discouraged from using such tools when appropriate.

Double-Blind Submission

ISE'26 will follow a double-blind review process. All submitted papers should conceal the identity of the authors. Both author names and affiliations must be omitted. In addition, the following rules should be addressed:

After the paper's acceptance, all the paper information (without anonymization) can be included in the camera-ready version.

Any questions about the preparation of the paper following the double-blind rules can be sent to the Program Committee Chairs.

Organization, Chairs and Important Dates

Organization:

ISE is a workshop organized and promoted by UFCG, UEG, and supported by VIRTUS/UFCG.

Program Committee Chairs:

ISE Steering Committee:

Important Dates:

Definitive

Keynote

Person keynote picture profile

Evandro Costa

Full Professor at UFAL

Title: Intelligent Agents in Software Engineering: From Classical Rationality to LLM-based Agents and Neuro-Symbolic Integration

Description: This keynote offers a structured perspective on the evolving relationship between intelligent agent research and Software Engineering, spanning classical AI foundations, LLM-based agents, and the current frontier of Neuro-Symbolic Architectures. Beginning with the rational agent model and its implications for software system design, the talk examines Agent-Oriented Software Engineering as an early attempt to make agent abstractions first-class citizens of the SE process. It then addresses Multi-Agent Systems, their coordination mechanisms, and frameworks such as JADE and SPADE. The second half of the talk focuses on the paradigm shift introduced by LLM-based agents, analyzing the ReAct pattern and its continuity with classical BDI architectures, before turning to Neuro-Symbolic Architectures as a response to the verification and reliability challenges that purely neural agents cannot address. The keynote concludes with Neuro-Symbolic Architectures as a promising convergence point, and an assessment of what Software Engineering research must still deliver.

Short Bio: Evandro de Barros Costa is a Full Professor in the Computing Institute at Federal University of Alagoas (UFAL), Brazil. He received a B.S. degree in Computer Science from Federal University of Paraiba (UFPB), Brazil in 1988, an M.Sc. and a Ph.D degree in Electrical Engineering from UFPB in 1989 and 1997. His research interests include Generative Artificial Intelligence, agent and multi-systems, knowledge representation and reasoning, explainability in AI, machine learning, and data science. He has published more than 200 papers in international conferences and journals. He has been on the technical program committee of many conferences, such as IJCAI, AAMAS, UAI, AIED, ECAI, and has been reviewer on some international journals, such as Information Science, IEEE Transactions on Education, IEEE Transactions on Learning Technologies. Over the years, he has worked extensively on designing and developing AI for education and for fashion domain.

Person keynote picture profile

André Araújo

Senior Staff Research Scientist / Tech Lead Manager at Google DeepMind

Title: What’s missing in multimodal AI? Towards spatial awareness, effective tool use and fine-grained understanding

Description: The recent progress in multimodal AI systems has enabled a variety of applications, such as natural chatbot interactions, smooth information-seeking journeys and promising robotics deployments. Despite the significant advancements in the past few years, these systems still face challenges in a number of areas. In this talk, I will highlight recent work from my team and collaborators to enhance three important aspects: spatial awareness, effective tool use and fine-grained understanding. Taken together, these ideas offer directions for designing ever-more-capable multimodal systems and experiences.

Short Bio: André Araujo is a Senior Staff Research Scientist / Tech Lead Manager at Google DeepMind, based in the Sao Paulo (Brazil) office. His current work focuses on multimodal AI (computer vision, machine learning and related areas). André leads a team of researchers distributed between Mountain View (US), Sao Paulo (Brazil) and Zurich (Switzerland). He graduated with a Ph.D. from Stanford University in 2016 and is a Senior Member of the IEEE.

Awards

Distinguished Paper

The ISE/CBSoft Steering Committee will appoint a committee to award the best papers of ISE Technical Papers. Awarded papers will be announced during the ISE/CBSoft. The best papers will be invited to submit an extended version to the Journal of Software Engineering Research and Development (JSERD).

Outstanding Reviewer

The ISE/CBSoft Steering Committee values the important role played by Program Committee members in ensuring the quality of the event. As a way to recognize their commitment, the organizing committee will present an Outstanding Reviewer certificate to individuals who demonstrate excellence in reviewing, through timely feedback, constructive and well-structured evaluations, and active engagement in committee discussions.

ISE Program Committee

Program

Presentation Format

Each accepted paper will be allocated 10 minutes for presentation, followed by 3 minutes for audience interaction (questions, suggestions, and comments). Authors are encouraged to use slides or any other presentation format they find suitable (such as prototype demonstrations or videos), as long as the 10-minute time limit for the presentation is respected.

Paper Publication

Accepted papers will be assigned a DOI and published in the event proceedings, which will be made available online through the SBC OpenLib, the SBC's content portal, in the Anais do ISE/CBSoft series, accessible at: https://sol.sbc.org.br/index.php/ise/index.

Workshop Program

To appear

Registration and Participation

At least one paying author of each accepted paper must register for the paper to be included in the event proceedings. Authors with more than one accepted paper, in any CBsoft event, may pay a single registration fee plus an “additional publication fee” for each extra paper. The ISE/CBSoft Steering Committee reserves the right to exclude from the proceedings any paper that is not presented during the event.

Previous Editions

ISE 2021 ISE 2022 ISE 2023 ISE 2025
Logo VIRTUS UFCG

ISE'26 is a workshop organized and promoted by UFCG, UEG, and supported by VIRTUS/UFCG.

Copyright 2026© - VIRTUS/UFCG