Annals of Computer Science and Intelligence Systems, Volume 48
Position Papers of the 21st Conference on Computer Science and Intelligence Systems (FedCSIS)
Preface
Dear it is our pleasure to present to you the Position Papers of the 21th Conference on Computer Science and Intelligence Systems (FedCSIS 2026), which took place on 23-26 August, 2026, in Riga, Latvia.
Position papers comprise two categories of contributions -- challenge papers and emerging research papers. Challenge papers propose and describe research challenges in theory, or practice, of computer science and intelligence systems. Papers in this category are based on deep understanding of existing research or industrial problems. Based on such understanding and experience, they define new exciting research directions and show why these directions are crucial to the society at large. Emerging research papers present preliminary research results from work-in-progress, based on sound scientific approach but presenting work not completely validated as yet. They describe precisely the research problem and its rationale. They also define the intended future work including the expected benefits from solution to the tackled problem. Subsequently, they may be more conceptual than experimental.
FedCSIS 2026 was chaired by Oksana Nikiforova. Moreover, Andrejs Romānovs was the Chair of the Organizing Committee. Furthermore, this year, FedCSIS was organized by the Polish Information Processing Society (Mazovia Chapter), IEEE Poland Section Computer Society Chapter, Systems Research Institute Polish Academy of Sciences, The Faculty of Mathematics and Information Science of the Warsaw University of Technology, The Faculty of Electrical and Computer Engineering of the Rzeszów University of Technology, The Faculty of Electrical Engineering af the Riga Technical University and The Faculty of Engineering and Management of the University of Technology and Arts in Applied Sciences.
FedCSIS 2026 was technically co-sponsored by IEEE Poland Section, IEEE Serbia and Montenegro Section, IEEE Estonia Section, IEEE Latvia Section, IEEE Lithuania Section, IEEE South Africa Section, Serbia and Montenegro Section Computational Intelligence Society Chapter, IEEE Young Professionals Affinity Group, Novi Sad, Republic of Serbia, IEEE Romania Section WIE Affinity Group, Czechoslovakia Section Computer Society Chapter, Poland Section Computer Society (Gdańsk) Chapter, Estonia Section Computer Society Chapter, Latvia Section Computer Society Chapter, Lithuania Section Computer Society Chapter,
FedCSIS 2026 was sponsored by InfoSoftware Polska as well as MDPI Information, MDPI Electronics, MDPI Technologies, MDPI Applied Sciences journals. Moreover, FedCSIS 2026 has been conducted under patronages of: INFORMS Poland Chapter, Latvian Information and communications technology association (LIKTA), Naukowe Towarzystwo Informatyki Ekonomicznej, International Rough Set Society, Web Intelligence Consortium, Polish Neural Network Society, European Society for Fuzzy Logic and Technology, Polish Artificial Intelligence Society (PSSI). Finally, media patronage was provided by Forum Akademickie.
During FedCSIS 2026 four keynote speakers delivered lectures providing a broader context for the conference participants. These presentations were:
- Ginter, Filip, TurkuNLP, Department of Computing University of Turku, Finland and ELLIS Institute, Finland
Keynote title: The Challenges of Tracing Meaning at Scale across Multilingual Historical Texts - Pechenizkiy, Mykola, Department of Mathematics and Computer Science, Eindhoven University of Technology, Netherlands
Keynote title: Advancing the Science of Safe and Trustworthy AI Evaluation - Prodan, Radu, Edge AI at the Department of Computer Science, University of Innsbruck, Austria
Keynote title: Distribute and Learn 1 Billion - Sabou, Marta, Vienna University of Economics and Business (WU), Austria
Keynote title: Neuro-Symbolic Knowledge Engineering: Emerging Paradigms and Challenges
Moreover, four past FedCSIS keynote speakers have been invited to prepare and deliver special contributions, which refer to the core focus of the conference series. These were:
- Atiquzzaman, Mohammed, School of Computer Science, University of Oklahoma, USA
EnsembleClassifier: Using Machine Learning to Select Exact Packet Classification Structures for Partitioned Rule Subsets - Blum, Christian, Artificial Intelligence Research Institute (IIIA-CSIC), Spain
Contribution title:Preserving Optimization Algorithm Expertise by means of Executable Algorithm Knowledge Graphs: A Worked Example on the TSP - Skowron, Andrzej, Systems Research Institute Polish Academy of Sciences, Poland
Contribution title:Evolution of Approximation Spaces in Rough Sets: From Relational Systems to Complex Adaptive Systems - van der Aalst, Wil M.P., RWTH Aachen University, Germany
Contribution title: Mining Bureaucracy Debt: Process Intelligence for Evidence-Based Deregulation
At the time, when you are reading this text, videos of the keynote presentations and of invited contributions, delivered during the FedCSIS 2026 conference, are already available on the official conference website (www.fedcsis.org). We warmly encourage you to visit the website and watch these recordings to gain additional insights and perspectives shared by distinguished speakers.
This year, the structure of the conference remained the same as last year. FedCSIS 2026 had a single Main Track with 5 Topical Areas, which was supported by 16 Thematic Sessions and an Industrial Track. Such a structure emphasizes the integrity of the conference and its closeness to the issues that are crucial for the world around us. Here, we recognize the fact that today (in 2026) it is difficult to envision research (and its applications) without an intelligence component. Reflecting this, all five Topical Areas, which constitute the Main Track, represent various aspects of Intelligence Systems. Moreover, the Thematic Sessions provide further insights into selected areas of Intelligence Systems, approached from different perspectives.
Articles are ordered alphabetically by the last name of the first author. The specific Topical Area or Thematic Session that given contribution was associated with is listed in the article metadata. Each contribution, found in this volume, was refereed by at least two referees.
Delivery of FedCSIS 2026 required a dedicated effort of many people. We would like to express our warmest gratitude to all Topical Area Curators, Thematic Session organizers, members of the FedCSIS 2026 Senior Program Committee, and members of the FedCSIS 2026 Program Committee (a total of more than 300 individuals), for their hard work in attracting and reviewing all submissions. We thank the authors of papers for their great contribution to the theory and practice of Computer Science and Intelligence Systems. We are grateful to Keynote and Invited Speakers for sharing their knowledge and experiences with the participants. Last, but not least, we acknowledge Andrejs Romānovs, Megija Krista Miļūne, Rihards Bobkovs, Uldis Karlovs-Karlovskis, Kristaps Babris, Veronika Tatjana Vucāne, Zane Bičevska Anastasiya Danilenka and Paweł Szmeja, as well as a fantastic group of student helpers. We are very grateful for your efforts!
We also hope to meet you again for the 22nd Conference on Computer Science and Intelligence Systems (FedCSIS 2027), which will take place in Gdynia, Poland, on September 12-15, 2027.
Co-Chairs of the FedCSIS Conference Series
Marek Bolanowski, Rzeszow University of Technology, Poland.
Maria Ganzha, Warsaw University of Technology, Poland and Systems Research Institute Polish Academy of Sciences, Warsaw, Poland.
Marek Grzegorowski, Samsung Research, Warsaw, Poland.
Leszek Maciaszek, Wrocław University of Economics and Business, Wrocław, Poland and Macquarie University, Sydney, Australia.
Marcin Paprzycki, Systems Research Institute Polish Academy of Sciences, Warsaw Poland and Management Academy, Warsaw, Poland.
Andrzej Paszkiewicz, Rzeszów University of Technology, Poland.
Dominik Ślęzak, Institute of Informatics, University of Warsaw, Poland and QED Software, Poland and DeepSeas, USA.
Position Papers
-
Designing Industrial Interfaces for Digitally and Functionally Illiterate Users: A Case Study in a Manufacturing Environment
3200 Digital transformation, Human-machine interface, Digital Illiteracy, Functional Illiteracy, User-centered design Felipe Dias Abrahão, Leandro Durães, Arthur Pedro Fortes Silva, Vanessa Tavares de Oliveira Barros, Rodolfo Miranda de Barros, pages 1–8. ISM -
Improved Estimation of Soil-to-Sugar Beet Ratio Using Deep Learning and Domain Adaptation for Precision Agriculture
2778 Domain adaptation, Agriculture automation, Soil estimation, Sugar beet, Image segmentation, Deature extraction Elham Alfuqara, Matthias Hien, Raphael Roeser-Mueller, Johann Brunner, Tobias Dobmeier, pages 9–17. AgriAI -
Why Federated, Privacy-Preserving Data Spaces are the Future of Digital Pathology
8271 Federated Data Space, Privacy-Preserving Data Spaces, Digital Pathology, AI in Healthcare, Bio Image Processing. Mohammadreza Azimi, Henrik Gabrielyan, pages 19–24. DSH -
Towards Intelligent Incident Response: A Review of Adaptive and Automated Planning and Orchestration Approaches
5303 Cybersecurity, Incident Response, Security Orchestration Abir Benhizia, Abir Laraba, Mohamed Ali Kandi, Romain Laborde, pages 25–30. NEMESIS -
Implementing Communities of Practice with Social CRM
9749 Communities of Practice, Social CRM, digital platforms Cristina Alejandra Barahona Cabrera, Olaf Reinhold, pages 31–38. DSH -
Tacit Knowledge and Hermeneutics: Critical Reflections on Automated Text Interpretation with Large Language Models
5725 Automated Text Interpretation, LLMs, Tacit Knowledge Jens Dörpinghaus, Michael Tiemann, pages 39–48. AI‑HuSo -
Ontology-grounded logic reasoning for validating structured LLM outputs
4410 LLM output validation, Answer Set Programming, formal methods, three-way decision, ontology-based validation András Földvári, András Pataricza, pages 49–56. MDASD -
Toward Neurocognitive Intelligence for Assistive Humanoid Robots in Geriatric Care
0447 Elderly care assistance, humanoid systems, biological cognition, human-robot-interaction Noorbakhsh Amiri Golilarz, Jiacheng Li, Shahram Rahimi, Andy Perkins, pages 57–64. AAIA -
Physics-Based State-of-Charge Modelling and Operational Optimization for a Battery-Electric Shunting Locomotive Digital Twin
1984 digital twin, battery locomotive, lead-acid battery, state of charge, Ampere-hour integration, Peukert effect, recharge optimization, route simulation, railway telemetry Tsvetomir Gotsov, Venelin Todorov, pages 65–70. CO -
A Hybrid GraphRAG Approach to Generating Explainable Recommendations
3730 GraphRAG, knowledge graphs, large language models, recommendation systems, explainability Francisco Intriago-Usca, Janneth Chicaiza, pages 71–78. ISM -
Towards Inclusive Intelligence: Defining the Research Challenges at the Intersection of Artificial Intelligence, Universal Design for Learning, and Workplace Training
1991 Artificial intelligence, Universal Design for Learning, Workplace learning, Accessibility, UDL-Informed Evaluation Framework for AI-Powered Corporate, Workplace Training, Learning and development, Generative AI, AI in learning and development Aina Irbe, Eduardo Meza Etienne, pages 79–86. EDUC‑AI‑TION -
Browser-Deployed Lightweight YOLO for Banana Ripeness Classification on Commodity Edge Devices
5575 banana ripeness classification, ONNX runtime web, browser deployment, YOLO, edge inference, leave-one-day-out cross-validation Sumetee Jirapattarasakul, Sirinya Thanyacharoen, Giuseppe Riccardo Leone, Thaweesak Yingthawornsuk, pages 87–92. AgriAI -
Engineering Least-Privilege Agentic AI Workflows: Patterns, Anti-Patterns, and Practitioner Guidance
6361 agentic AI, least privilege, tool-call security, MCP, policy-as-code, multi-agent systems, ephemeral credentials, OPA Rajat Khanna, Jatin Nandal, pages 93–97. INDT -
Multi-Candidate Synthesis in Multimodal Machine Translation
4324 Multimodal Machine Translation, Large Language Models, Machine Translation, Visual Language Models Jan Kolwicz, Szymon Niewiadomski, Kamil Guttmann, Michał Ciesiółka, Krzysztof Jassem, pages 99–103. CNLPS -
A Breakdown of Feature Importance Ranking and Trimming (FIRT) for Maternal Health Risk Prediction Using Physiological Factors and Machine Learning
2543 Maternal health risks, machine learning, dimensionality reduction, feature selection, feature ranking, RFE, KNN, SVC, ANOVA F-test Shilpa Mahajan, Garima Sharma, Abdullah Ayub Khan, Asif Ali Laghari, Angela Maria. Alves, Gabriel Gomes de Oliveira, Vania Vieira Estrela, pages 105–112. DSH -
Mapping data centre energy consumption to carbon-dioxide emissions
9143 data centre, renewable energy, energy grid, digital twin Julian Meyer, Alexander Kammeyer, Florian Burger, Daniel Lübbert, Katinka Wolter, pages 113–118. IDPI -
Operationalising Governance for AI-Generated Content: A Socio-Technical Approach Based on the MATH–COPE Framework
8932 AI governance, generative AI, copyright, training data, provenance, AI regulation, system-level governance, data pipelines Anja Maria Neubauer, pages 119–125. AI‑HuSo -
Challenges in Modeling Historical Language Style for Diachronic Style Recognition
9713 diachronic NLP, historical language modeling, Transformer Encoder, FiLM, style recognition, historical corpora, SentencePiece tokenizer Adrian Niedziółka-Domański, Jarosław Bylina, pages 127–133. CNLPS -
Challenges in Stabilizing Training for Diachronic Text Style Transfer
6104 diachronic text style transfer, CycleGAN, Transformer, adversarial training, Gumbel-Softmax, FiLM conditioning, autoregressive text generation, training stabilization Adrian Niedziółka-Domański, pages 135–141. CANA -
StructRL: Recovering Dynamic Programming Structure from Learning Dynamics in Distributional Reinforcement Learning
6447 Reinforcement Learning, Dynamic Programming Structure, Sample-Efficient Learning Ivo Nowak, pages 143–148. AAIA -
Cross-Dataset Component Ablation of a Hybrid Model for Robust Beat-Level ECG Arrhythmia Classification
6565 ECG arrhythmia classification, cross-dataset robustness, domain generalization, feature ablation Jonas Mindaugas Rimšelis, Viktor Medvedev, Povilas Treigys, Žygimantas Abramikas, Jurgita Markevičiūtė, Jolita Bernatavičienė, pages 149–155. DSH -
Regularizing the Prompt: Stable Model-Agnostic LLM Responses Through Inversion Theory
1450 LLM, prompt engineering, model-agnostic, geophysical inversion, Tikhonov regularization, CACE Saulo Silva, pages 157–164. INDT -
Re-evaluating Polish LLMs in RAG: The Evolution of Bielik and PLLuM and the Limits of Open-Source Evaluators
9874 Retrieval-Augmented Generation (RAG), large language models (LLMs), LLM-as-a-Judge, evaluation Marcin Szewczyk, Tymon Pluciński, Krzysztof Maksim, Szymon Bartanowicz, Krzysztof Jassem, pages 165–169. CNLPS -
Universal Design and User-Centered Design: A UX Methodology Applied to the Democratization of Telesservices for People with Disabilities
4197 User-Centered Design, Universal Design, UX Methodology, Digital Accessibility, Telemedicine Laura Tejada Vergara, Felipe Dias Abrahão, Rodolfo Miranda de Barros, Vanessa Tavares de Oliveira Barros, pages 171–178. ISM -
UX and Digital Accessibility Maturity Diagnosis in the Brazilian Technology Market: A Survey Study with Design and IT Professionals
2446 UX Maturity, Digital Accessibility, User-Centered Design, Brazilian Technology Market Laura Tejada Vergara, Rodolfo Miranda de Barros, Vanessa Tavares de Oliveira Barros, pages 179–189. ISM -
Backtesting Split Frequency–Severity Models in Motor third-party liability (MTPL) Insurance Under Claims Inflation
3531 MTPL, claims inflation, frequency-severity model, backtesting, threshold migration, non-life insurance Ondřej Vít, Lubomír Seif, Lubomír Štěpánek, pages 191–196. DSH -
Anomaly Detection with a Blockchain Audit Trail for Open Banking Anti-Fraud: Prototype Evidence and Research Agenda
1099 Open Banking, fraud detection, anomaly detection, blockchain audit trail, PSD2, financial compliance Krzysztof Piech, pages 197–204. ISM -
Trusted or Tricked? A Multi-Layer Robustness Scoring Framework for Evaluating LLMs Against Public Health Misinformation
3733 Artificial Intelligence, Centers for Disease Control and Prevention, Large Language Models, Public Health, Misinformation Razan El Mais, Nourhan Hamze, Chaymaa Abbass, Hadi Al Mubasher, Mariette Awad, pages 205–212. DSH
E-mail:
Phone/fax: +48-89-5246089