The LABORIS - Testing Systems Research Group is a specialized research unit operating within the Center for Innovation in Engineering and Industrial Technology (CIETI) at the Instituto Superior de Engenharia do Porto (ISEP/P.Porto). The group focuses on research on remote laboratories, engineering education, and design for debug and test. By bridging the gap between hardware testing and pedagogical design, LABORIS drives innovation in how engineering concepts can be taught and practically applied through remote-access technologies. The ongoing research and development focus on the use of standard web browsers to control actual physical equipment housed at partner institutions worldwide, instead of relying on computer simulations or being restricted to physical campus benches.
Simultaneously the group is managing the SHOREWINNER Project (Southern European Community for Offshore Wind Energy) a co-funded European Union initiative designed to build a specialized workforce for the rapidly growing offshore renewable energy sector. This four-year initiative (2024–2028) brings together a consortium of 22 partners across five Southern European countries: Portugal, Spain, Italy, Greece, and Cyprus.
The primary mission of the project is to bridge the gap between labor market demands and training by establishing five Centres of Vocational Excellence (CoVEs). These specialized centers cooperate locally with industry representatives, academic bodies, and vocational schools to design targeted, high-quality educational programs that address regional energy goals.
==> Abstract (PDF)Context. Modern software systems rely heavily on dependencies distributed through Maven Central, and vulnerabilities frequently reside in transitive rather than direct dependencies [1]. Tools such as OWASP Dependency-Check are effective at flagging these known vulnerabilities, yet a large share of the flagged code is never actually executed by the depending application [2].
Problem. Because such tools operate without this context, developers face a binary choice whenever a vulnerability is reported: upgrade immediately and risk breaking the build — a risk that materialises in a substantial share of Maven releases [3] — or postpone the update and accept the security exposure.
Objective. This work designs, implements and evaluates a Contextualised Risk-Based Framework (CRBF) that supports Maven dependency-update decisions by combining exploitability, reachability, ecosystem stability and compatibility signals into a single risk score, and by selecting an optimal remediation plan under a configurable effort budget.
Method. CRBF is implemented as a Maven plugin with a hexagonal architecture. For every dependency, known vulnerabilities are retrieved from the OSV database and enriched with EPSS exploitation scores; a static call graph built with WALA determines whether the vulnerable code is reachable from the project's own classes; binary compatibility between the current and candidate fix versions is assessed with JApiCmp; and ecosystem adoption and maintenance indicators are retrieved from the Goblin Weaver dependency graph [4]. These signals are aggregated into a contextual risk score, and a Z3 SMT solver selects the subset of upgrades that maximises risk reduction within the effort budget, following the constraint-based approach to dependency reasoning explored in prior work [5].
Results. CRBF has been implemented and exercised end-to-end on individual Maven projects. A comparative evaluation against OSV-Scanner, with and without call-graph analysis, across a broader sample of open-source Maven projects is currently under way, and its outcomes will be presented at the symposium.
Conclusions. By turning severity, exploitability, reachability, stability and compatibility into a single, effort-bounded recommendation, CRBF extends the balance between security and update cost explored in prior dependency-update research [6] and offers developers actionable guidance, addressing the low update rates that non-contextual vulnerability reports have been shown to produce [7].
References
[1] C. Nachuma, M. M. Hossan, A. K. Turzo, and M. Zibran, “Decoding Dependency Risks: A Quantitative Study of Vulnerabilities in the Maven Ecosystem,” in Proc. IEEE/ACM Int. Conf. Mining Software Repositories (MSR), 2025, pp. 270–274, doi: 10.1109/MSR66628.2025.00048.
[2] M. Keshani, S. Vos, and S. Proksch, “On the relation of method popularity to breaking changes in the Maven ecosystem,” Journal of Systems and Software, vol. 203, p. 111738, 2023, doi: 10.1016/j.jss.2023.111738.
[3] S. Raemaekers, A. van Deursen, and J. Visser, “Semantic versioning and impact of breaking changes in the Maven repository,” Journal of Systems and Software, vol. 129, pp. 140–158, 2017, doi: 10.1016/j.jss.2016.04.008.
[4] D. Jaime, J. El Haddad, and P. Poizat, “Goblin: A Framework for Enriching and Querying the Maven Central Dependency Graph,” in Proc. IEEE/ACM Int. Conf. Mining Software Repositories (MSR), 2024, pp. 37–41.
[5] A. G. Márquez, Á. J. Varela-Vaca, M. T. Gómez López, J. A. Galindo, and D. Benavides, “Depex: A software for analysing and reasoning about vulnerabilities in software project dependencies,” SoftwareX, vol. 30, p. 102152, 2025, doi: 10.1016/j.softx.2025.102152.
[6] D. Jaime, P. Poizat, J. El Haddad, and T. Degueule, “Balancing the Quality and Cost of Updating Dependencies,” 2025. [Online]. Available: https://hal.science/hal-04684254v1
[7] K. Huang et al., “Characterizing usages, updates and risks of third-party libraries in Java projects,” Empirical Software Engineering, vol. 27, no. 4, 2022.
The source code is openly available at https://github.com/CRBF-Project/risk-based-framework.
==> Abstract (PDF)A systematic literature review is presented on Machine Learning (ML) applications for predicting sports outcomes, following PRISMA 2020 guidelines, analyzing 30 primary studies from 2020-2025 sourced from Scopus and Google Scholar. Key findings highlight football and basketball as dominant sports (~67% of studies).
Artificial Neural Networks (~30%) is the most commonly used method followed by tree-based methods (~24%), statistical & regression methods (~17%) and instant-based and kernel methods (~15%). Most methods use feature selection (~81%), over formal dimensionality reduction, and accuracy is the primary metric (80% of studies). Prediction rates vary from 55% to 93%.
==> Abstract (PDF)Semi-autonomous robot assistants can support both teaching and administrative activities in higher education, but existing research often examines individual robot applications in isolation. This fragmentation makes it difficult to determine how different robot-assisted functions can be combined, what level of autonomy is appropriate, and which implementation conditions are required for reliable classroom use. This study addresses the problem by synthesizing lessons from previously reported robot-assisted educational scenarios into a preliminary set of design considerations for semi-autonomous robot assistance in higher education.
The analysis draws on prior literature and classroom implementations involving attendance checking, question answering and guidance, practicaltask pre-evaluation, and immediate feedback collection. The resulting design considerations cover the placement of robot support within the teaching process, the division of responsibilities between the robot and the lecturer, the relationship between robot functions and supporting software services, and implementation factors including usability, reliability, privacy, ethics, and institutional support.
Results from the completed classroom studies show generally positive acceptance of robot-assisted support for routine administrative and structured instructional tasks, while also demonstrating that acceptance depends on reliable operation, appropriate interaction modes, lecturer control, and clear practical value. The main contribution is a cross-scenario synthesis of design considerations that moves beyond isolated robot use cases and identifies the elements that should be explicitly addressed when planning semi-autonomous robot assistance in higher education. These considerations provide the basis for a second structured feedback round with researchers and practitioners to assess completeness, clarity, and practical relevance and to guide further refinement.
The widespread adoption of generative artificial intelligence (GenAI) tools is changing academic writing in higher education and, consequently, some of the assumptions traditionally associated with its assessment. This presentation draws on teaching experience in supervising and assessing undergraduate Project/Internship reports and Master’s dissertations in Computer Engineering, considering changes observed following the widespread availability of GenAI tools.
Before the emergence of these technologies, academic written work frequently showed grammatical and syntactic errors, difficulties in structuring technical discourse, limited vocabulary, and problems with clarity and coherence. These aspects often required considerable attention during assessment and oral examinations and could significantly affect the evaluation of the written work.
More recently, a noticeable change has emerged. Reports and dissertations tend to be linguistically more polished, better structured, and stylistically more sophisticated. While this represents a positive development in terms of written communication, it also raises an important assessment challenge: the formal quality of academic writing may no longer provide the same evidence about a student’s ability to independently structure, articulate, and critically communicate technical knowledge.
Rather than treating GenAI solely as a threat to academic integrity, this presentation considers its growing use as an opportunity to rethink assessment practices. Greater emphasis may need to be placed on students’ understanding of their own work, justification of technical decisions, critical thinking, implementation skills, and ability to defend their choices during oral examination.
The presentation discusses these changes from a teaching perspective and considers how assessment in engineering education might evolve when producing polished academic text becomes increasingly accessible through AI-assisted writing.
Preparing higher education students for an increasingly global labour market requires more than disciplinary knowledge. It requires opportunities to develop transversal competences, international awareness and the ability to collaborate across cultural and disciplinary boundaries. Since 2008, the BlendEd initiative has explored how innovative learning environments can integrate these dimensions into regular higher education without requiring major curricular restructuring or extensive additional resources.
This paper presents an 18-year evolution of this pedagogical work, from the initial Multinational Undergraduate Team Work (MUTW) project to the current Africa BlendEd Course (ABC BlendEd) initiative. Across successive editions, the model evolved from homogeneous, teacher-defined projects to international and multidisciplinary teams addressing authentic challenges proposed by companies, with increasing integration of blended mobility, entrepreneurship, intercultural learning and cooperation with external stakeholders. More than 500 students and over 30 companies have participated in the European implementation, while the approach has progressively expanded beyond Europe.
The paper analyses the main pedagogical developments, the rationale behind each evolutionary step, and the lessons emerging from sustained implementation. ABC BlendEd represents the latest stage of this process, adapting the accumulated experience to African higher education and agriculture. The experience suggests that international, challenge-based and multidisciplinary project learning can provide a practical pathway for embedding employability development within existing curricula while simultaneously widening access to meaningful international exposure.
Permutable lists, sets, and multisets can be looked at in two ways: as unordered collections, whose entropy is usually captured by treating them as opaque binary strings once their order is set aside, or as sequences, whose order is encoded separately from the elements it arranges. Between these two views lies an open question — whether the order in which elements appear can itself be read as information, and folded back into the representation of the data rather than discarded or kept apart from it.
This dissertation takes up that question, proposing a recursive order-based compression method that bridges sequence representations and multiset entropy theory, in which a multiset is represented as a list of its original, or compressed, elements. At the heart of the method is a new permutation encoder that lets the order of a list's elements carry part of the list's own information; this order-derived information is then recursively unfolded into further elements, each sublist read against the order of the one that came before it. This recursive folding is shown to yield savings that are asymptotically optimal, converging to the entropy of the underlying permutation as the list grows, all while the resulting representation remains directly indexable, with random access preserved throughout. The method is realized through a Fenwick tree for O(log K) rank queries, together with worst-case and average-case partitioning strategies, brought together in an end-toend encoding and decoding pipeline, and its practical compression behavior is characterized experimentally.
The resulting contribution is a general encoding scheme that draws order-based and multiset compression into a single, coherent framework — a representation that is both compact and immediately usable, without giving up one quality for the other.
Effective cyberattack detection requires more than collecting logs and deploying monitoring tools. Security teams need to translate relevant threat scenarios into observable events, identify suitable data sources, and maintain detection rules as systems and attacker behaviour change. Detection engineering provides a systematic approach to connecting these tasks and supporting continuous improvement in security operations.
This presentation introduces the detection engineering lifecycle, from the identification of detection requirements and suitable telemetry to the development, testing, and refinement of detection rules. Using Wazuh as a practical example of a security information and event management (SIEM) platform, it examines how agents collect events, how decoders extract relevant fields from log data, and how rules use this information to generate security alerts. Particular attention is given to the relationship between threat scenarios, available telemetry, and detection logic, as well as the distinction between an observed event, an alert, and a confirmed security incident. Feedback from security operations teams is considered an essential input for reducing false positives and adapting detections to operational requirements.
The presentation also explores the role of artificial intelligence (AI) in supporting detection engineering, including assistance with decoder and rule configuration and the enrichment of alerts with additional context. These applications are discussed alongside their limitations, including unreliable assessments, the exposure of sensitive data, and prompt injection through untrusted input. The discussion emphasises the importance of validating generated configurations and checking automated assessments against the available evidence before using them to inform security decisions.
By connecting threat analysis, technical implementation, and operational feedback, the presentation provides a structured introduction to developing and maintaining detection capabilities. It establishes a basis for understanding the detection engineering lifecycle and assessing where AI assistance can contribute to security operations and where human expertise remains essential.
The education of Deaf and hard of hearing individuals is critically hindered by a global shortage of qualified sign language (SL) teachers and a lack of accessible, personalized learning tools. This paper presents VISTA-SL, an innovative project that leverages advanced technology to address these challenges and promote the fundamental rights, dignity, and unique culture of the Deaf community.
The project, implemented within the Erasmus+ programme, aims to develop a comprehensive SL teaching ecosystem. Its core objectives are to mitigate the teacher scarcity through the creation of a technology-enhanced teaching toolkit. This toolkit integrates real-time visual feedback using Augmented Reality (AR) and avatars, alongside a Large Language Model (LLM)-based virtual instructor for course personalization and coaching. To foster inclusion and diversity, the project is concurrently developing interactive curricula and educational games in four European sign languages. These resources will be disseminated via a low-cost, accessible e-learning platform.
The implementation methodology encompasses the development of all software tools, the co-creation of pedagogical materials with SL linguists and educators, transnational consortium meetings, training activities, and wide-reaching dissemination through multiplier events. Expected outcomes include open-source software tools and detailed technical reports, interactive learning curricula, and comprehensive evaluation reports from diverse user groups. By integrating explainable visual feedback and machine learning, VISTA-SL aims to establish a new standard for SL education, enhance its quality and effectiveness, and significantly contribute to the inclusion and empowerment of Deaf learners and educators across Europe.
==> Abstract (PDF)City Regions and Businesses boosting Nature Positivity (KYLKI) is a collaborative project of the city of Vantaa, Metropolia UAS, Laurea UAS, and the Finnish Environment Institute. The project members are working together to support the transition of businesses and cities in the metropolitan area towards nature-positive practices. The project aims to promote solutions that foster sustainable business, environmental responsibility, and help cities achieve their biodiversity goals.
It aims to establish a network, develop methods and tools for assessing and reducing the environmental impacts of businesses and cities, and test new innovations in urban settings. Through activities such as data collection, co-creation within an RDI framework, and pilot projects, the project will create opportunities for business innovation, strengthen the competitiveness of small and medium-sized enterprises (SMEs), and promote the development of environmentally responsible markets.
One of the work packages in the project deals with collecting environmental data to assess functionality of rain gardens built for managing stormwater runoff. There are lots of sensor and data loggers available from different manufacturers. Typically, data loggers are configurable to support sensors with compatible electrical interface, but the loggers are tied to the manufacturer’s cloud infrastructure which limits the applicability of the data or adds complexity to the data collection. To address these issues an open-source data logger was developed using publish-subscribe model. Transport layer and data encoding are separated from data collection letting user to choose both data encoding and cloud endpoint freely. The design even allows user to run multiple cloud senders to pass the data to multiple clients without adding an intermediate transport agent.
Context. Docker is the de facto standard for containerization, and Dockerfiles act as executable specifications that determine how images are built. Violations of recommended practices, known as Dockerfile smells, are pervasive in open-source projects [1] and degrade security, build performance, reproducibility and maintainability [2], [3].
Problem. Linters such as Hadolint report every violation at the same level of importance, while automated repair tools such as Parfum [4] and DockerCleaner [5] fix smells in isolation. Neither indicates which smell should be addressed first when several coexist, nor makes explicit that repairing one smell may degrade a different quality attribute: pinning a dependency version, for instance, improves reproducibility while adding maintenance burden. Since developers do not perceive all smells as equally urgent [6], this absence of prioritization support leaves refactoring decisions unguided.
Objective. To design, implement and evaluate a prioritization mechanism that ranks Dockerfile repairs by their expected impact on four quality attributes, while making the trade-offs between those attributes explicit rather than hidden inside a single score.
Method. Following Design Science Research [7], a literature-derived impact matrix assigns each repair action a signed impact on security, performance, maintainability and reproducibility. A Weighted Sum Model combines these impacts with developer-defined weights to produce one priority score per detected smell. The model is implemented in Docker Prioritizer, a Python command-line tool that uses Parfum as a detection and repair oracle. It was evaluated through an automated build–repair–rebuild pipeline over 57 Dockerfiles from open-source repositories, with cold builds repeated three times per file and median values retained.
Results. Across the 24 Dockerfiles in which a smell was actually repaired, build time fell by 115 s and image size by 123 MB on average. On a held-out sample of 11 Dockerfiles not used for calibration, predicted and observed impacts agreed in direction for security (11/11), performance (10/11) and maintainability (9/11); reproducibility predictions proved least reliable.
Conclusions. Trade-off-aware prioritization of Dockerfile refactorings is feasible and yields measurable quality gains, while keeping every recommendation inspectable. The tool and the evaluation scripts are publicly available at https://github.com/delcioMonjane/dmei2.
References
[1] Y. Wu, Y. Zhang, T. Wang, and H. Wang, "Characterizing the Occurrence of Dockerfile Smells in Open-Source Software: An Empirical Study," IEEE Access, vol. 8, pp. 34127–34139, 2020.
[2] T. Durieux, "Empirical Study of the Docker Smells Impact on the Image Size," in Proc. ICSE 2024, pp. 2568–2579.
[3] G. Rosa, S. Scalabrino, G. Robles, and R. Oliveto, "Not all Dockerfile Smells are the Same: An Empirical Evaluation of Hadolint Writing Practices by Experts," in Proc. MSR 2024, pp. 231–241.
[4] T. Durieux, "Parfum: Detection and Automatic Repair of Dockerfile Smells," arXiv:2302.01707, 2023.
[5] Q.-C. Bui, M. Laukötter, and R. Scandariato, "DockerCleaner: Automatic Repair of Security Smells in Dockerfiles," in Proc. ICSME 2023, pp. 160–170.
[6] G. Rosa, F. Zappone, S. Scalabrino, and R. Oliveto, "Fixing Dockerfile smells: an empirical study," Empirical Software Engineering, vol. 29, no. 5, 2024.
[7] K. Peffers, T. Tuunanen, M. A. Rothenberger, and S. Chatterjee, "A Design Science Research Methodology for Information Systems Research," Journal of Management Information Systems, vol. 24, no. 3, pp. 45–77, 2007.
Artificial intelligence is changing how organizations operate, make decisions, and create new products and services. Its potential to improve productivity, support innovation, and use resources more efficiently has made AI an important part of many organizations' growth strategies. At the same time, the rapid adoption of AI has raised concerns about issues such as data privacy, fairness, transparency, accountability, security, and environmental impact.
This paper examines how organizations can use AI to support growth without losing sight of their broader responsibilities. It develops a conceptual framework that connects AI-driven growth with responsible business development. The framework includes six key areas:
strategy, responsible AI management, transparency, human involvement, sustainability, and ongoing risk management.
The paper argues that growth and responsibility should not be treated as separate or competing goals. When AI is used responsibly and guided by people, it can make work more efficient, encourage innovation, improve decision-making, and help use resources more sustainably. At the same time, responsible AI practices can strengthen trust, reduce organizational risks, and support long-term competitiveness. The paper concludes that the benefits of AI depend largely on how organizations choose to implement and govern the technology. Rather than viewing responsible AI as a constraint on innovation, organizations can treat it as an important capability for building sustainable growth and long-term value. The proposed framework provides a basis for further empirical research into how organizations can combine AI adoption with responsible and sustainable development
The convergence of Information Technology and Operational Technology in Industry 4.0 environments has significantly expanded the cybersecurity attack surface of industrial organisations. This paper presents findings from a cross-sectional survey conducted with 815 respondents – 224 operational technicians and 591 workers – in industrial and service-sector organisations, within the framework of the European Erasmus+ InCyS4.0 project.
Using purpose-built questionnaires, we assessed cybersecurity awareness, technical security measures, and risky user behaviours across different company sizes and sectors. Results reveal a significant gap between the security posture of large and small organisations, a low rate of formal cybersecurity training (26% of workers), and concerning risk behaviours such as use of personal email on company systems (42%) and transfer of personal files to organisational computers (40%). Among technicians, strengths were found in antivirus deployment (91.5%) and physical network protection (82.3%), while gaps remain in incident response planning (48.3%) and traffic analysis tooling (59.8%). These findings inform targeted training recommendations for industrial cybersecurity education.
In an ambitious move to revolutionize the European manufacturing landscape, a consortium of leading educational institutions, innovative small and medium-sized enterprises, and industry clusters from across Europe have joined forces to launch the "Digital Twin on Smart Manufacturing" project. This initiative has been made possible through the Erasmus+ programme, project number 101140156.
“Digital Twin” concept involves creating a virtual model of a physical object, process, or system. This digital representation is designed to accurately mirror the real-world counterpart in all its aspects, from physical properties and behaviours to processes and performance. The purpose is to allow for simulation, analysis, and control of the entity in a virtual environment, which can lead to insights and actions that improve the real-world counterpart's efficiency and maintenance.
The "Digital Twin on Smart Manufacturing" project aims to create a comprehensive European Digital Twin Skills ecosystem. The project will deliver an integral curriculum catering to the emerging professional needs of mid to high-level Information Technology and Operational Technology technicians across the European Qualifications Framework (EQF) Levels 4-8. This curriculum will be available in English and 5 other European languages, ensuring wide accessibility.
The primary objective of the project is to enhance the capacity of educational and vocational training providers to align their offerings with the dynamic requirements of the manufacturing industry. This is expected to result in several tangible outcomes, including:
To achieve these goals, the project will develop and disseminate a range of transformative educational resources and tools, some already available at https://digitaltwinproject.eu/, including:
The consortium comprises 2 VET providers, 1 HVET provider, 2 higher education institutions, 4 SMEs, 1 industry cluster, and 1 chamber of commerce from 5 European countries - Italy, Spain, Greece, Sweden and Bulgaria.
The project aims to facilitate fostering a skilled workforce equipped with the latest digital twin technologies, and thus make a significant contribution to the European Green Deal, the digital transformation of Europe, and the support of small businesses. The creation of Digital Twin occupation profiles and issuing micro-credentials will further solidify the project's long-term impact on the European labour market.
Scalable, interpretable, and intelligent network monitoring is essential for 5G and future networks. This talk introduces Aim5B, an AI-integrated semantic framework for 5G and beyond network management. Aim5B transforms unstructured logs from key 5G core network functions into a knowledge graph aligned with the semantic structure of control-plane events. Using a large language model (LLM), it translates natural language queries into Cypher graph queries, enabling precise log retrieval, event analysis, temporal correlation, and statistical summarization, without static parsing rules or predefined dashboards.
Deployed on a real private 5G testbed, Aim5B achieved full performance across precision, recall, F1, and Jaccard metrics for detailed, domain-specific queries across network function components, while reducing management traffic by approximately 99.9% through targeted, event-specific querying. We'll discuss how these results position Aim5B as an intelligence-driven approach to real-time, scalable network observability in future mobile systems.
Hello Robot’s Stretch 3 is “one-handed” innovative telescopic remote controllable service robot, studied thanks to open platform in many universities around the globe. It’s two-finger gripper, connected to agile wrist, is designed to be as universal as possible, and is capable of handling hard, soft and fairy heavy (around 2kg) objects. Still, there are application areas and conditions where universal gripper is not effective enough. In TalTech IT College, we have investigated under student projects and thesis work the usability of service robots in different environments. E.g. in laboratory environment, where even simple actions, such as handling a banana plug, can be a challenge to conventional robotic grippers. From there, the students have developed a gripper inspired by a bird’s beak, capable for greater force and precision, making it suitable for handling small socketed objects, buttons, rotating knobs etc. Another study is related to applicability of the Stretch platform e.g. in kindergarten playroom environment that challenge its sensors and processing capabilities till AI model deployment. The possible solutions were explored through semester projects and individual thesis.
Finally, the Stretch 3 robot was used at PhD-level to study of behavioural signals under remote control for operator identification and continuous authentication. Resulting research paper is at submission stage. There has been positive feedback from students who were eager to explore topics beyond of their regular curricula, such as RTOS, AI model development, materials science, data analysis, etc. The diversity of topics favours multidisciplinary teams.
At the same time, immaturity of open tools, despite generally good documentation and community support, remains a challenge. Addressing the complexity of both the problem and solution spaces requires a step-wise approach and dedicated time for experimentation. On the positive side, every successful solution paves a way for new challenges and learning opportunities.
The increasing availability of educational data, combined with the rapid advancement of artificial intelligence (AI) methods, is reshaping decision-making processes within higher education institutions. Data-driven approaches facilitate the systematic analysis of student-related data to support prediction, classification, and strategic management. Nevertheless, despite the growing body of research in this domain, the conceptualization of decision-making processes in higher education remains fragmented and insufficiently systematized.
The objective of this study is to develop a conceptual typology of decision-making domains in higher education, derived from a structured review of the extant literature. The analysis concentrates on research in educational data mining, learning analytics, and the application of AI in higher education contexts. The study employs a structured review of peer-reviewed publications retrieved from major academic databases. The literature is examined with the aim of identifying key analytical categories and organizing them into a coherent framework of decision-making areas. The findings result in the identification of core decision-making domains associated with academic processes, organizational management, and digital transformation. The proposed typology elucidates the interrelationships among these domains and reveals persistent gaps in the integration of data-driven approaches into institutional decision-making practices.
This study contributes to the conceptual development of higher education as a data-driven ecosystem and provides a foundation for subsequent empirical research focused on the prioritization of decision-making areas and the design and implementation of decision support systems.
==> Abstract (PDF)In the nowadays digital world many users are not able to understand the technical development and their results, but they are using them daily without taking over the responsibility for the decisions that are made on the basis of the tools, that are results of engineers’ decisions. Engineers are following the requests for technological support that are requested by their customers and are mostly not thinking about accountability and ethical decisions, because they are in the service of progress. With the rise of artificial intelligence, the complexity of the problem of responsibility and accountability has increased, what requires rapid support from the ethical point of view. The most effective would be courses that would educate and raise awareness in accordance with responsibility and ethics, but their development and integration into study program takes time, so it is necessary to propose what is already available. These are Professional Codes of Ethics, which may be mostly seen as rules of an association, but in reality, can ensure accountability and increase awareness.
In the paper, we will examine accountability with an emphasis on ethics, including cultural awareness and professional codes of ethics. These codes govern the behavior of individuals or groups of people within specific areas of the business environment. They establish and maintain the highest standards of professional conduct. Common features of such codes include avoiding conflicts of interest, breaches of confidentiality and privacy, and violations of laws, as well as ensuring knowledge of advanced technologies, the responsible use of information, and the maintenance of system integrity. They also involve applying fundamental ethical principles to professional activities in computing. However, Professional Codes of Ethics are not algorithms for solving ethical dilemmas. They serve as a foundation for ethical decision-making and as a framework for demonstrating professional responsibility and contributing to the public good.
==> Abstract (PDF)