Institutional Repository of Academic Research University of Macedonia
Not a member yet
2215 research outputs found
Sort by
Educators’ perceptions of the teaching of bioethics in primary and secondary schools in Greece
No full textABSTRACT: In recent years, a growing number of studies have emphasized the importance of integrating ethics education into the school curriculum. The objective of the present study is to investigate the perspectives of educators on the teaching of bioethics in primary and secondary education in Greece. A survey was conducted using google.forms. Through the official websites of the
Directorates of Primary and Secondary Education throughout Greece, the email addresses of
schools were identified. The link to the questionnaire was sent to these email addresses and the
director of each school was asked to forward it to his/her fellow teachers. A total of 1274 educators
responded to the survey. We found that there is strong support for integrating bioethics education
into both primary and secondary education, with a clear preference for incorporating it into existing courses rather than creating separate bioethics courses. The necessity of teaching bioethics is
viewed as slightly more important in secondary education compared to primary education, reflecting the increased complexity and relevance of bioethical issues as students mature. This is the first
large-scale survey in Greece to examine the views of teachers in primary and secondary schools on
issues related to bioethics. The study supports the argument for the introduction of bioethics education in the early years of school life. This, however, requires the enhancement of teacher training
programs in bioethics and the development of appropriate teaching materials and techniques for
use at each level of education.2512713
Identifying Inconsistent Temporal Triples in Temporal Knowledge Graphs
No full textKnowledge Graphs (KGs) are a popular means of representing information in a machine readable and interpretable way. This popularity, though, means that KG quality becomes increasingly important. However, most related research considers KGs as static, ignoring their evolutionary aspect. In this work, we focus on Temporal KG (TKG) quality and propose a novel method for detecting inconsistencies within a TKG’s triples by leveraging internal information. Our method involves automatically detecting all temporal relations of a TKG and their different variants, and by leveraging their support and confidence metrics, determining the dominant variant, while labeling as potentially inconsistent all triples following the other variants. We showcase the usability of our approach through a first case study on the YAGO Tiny KG, and discuss potential expansions.2676 CCIS410419New Trends in Database and Information System
Collaborative Uses of GenAI Tools in Project-Based Learning
No full textArtificial intelligence (AI) is forcing a dramatic transformation of the methods by which we acquire knowledge and engage in collaborative learning. Although there are several studies on how AI can support collaborative learning, there are no published studies examining how students can actually collaborate among themselves while interacting with AI tools. For this study, thirty postgraduate students were organized into teams of three, and each team developed a project mainly exploiting responses from ChatGPT, Google Gemini, and MS Copilot, as well as the internet and class resources. Each team selected a specific internet of things (IoT) application area and described the technologies and real-world cases in this area. Then, each team delivered a report with the full description of their project and their interactions with these generative AI (GenAI) tools and presented their work in class. Additionally, students answered an online questionnaire with closed- and open-ended questions and participated in focus group discussions. Members of each team collaborated to design prompts using five suggested modes of collaboration. Eventually, half of the students exploited all five collaborative modes, but they mostly liked and preferred three of these collaborative modes. On average, teammates initially disagreed 24% of the time but eventually reached an agreement. Students appreciated GenAI tools for their quick and well-structured responses, natural communication style, broad subject coverage, as well as their ability to simplify complex topics and support personalized learning. However, they expressed concerns about GenAI tools’ inaccurate and inconsistent responses and identified key risks, such as passive learning, over-dependence, outdated information, and privacy issues. Finally, students recommended that GenAI tools should provide a shared and well-organized discussion space for collaborative prompt asking, allowing all team members to simultaneously view each other’s prompts and the tool’s responses. They also advised source verification and proper training to ensure these tools remain supplementary rather than primary learning resources.15335
Open Access Academic Publications: Benefits and Risks
No full textThe presentation focuses on both the benefits and the risks that emerge during
the process of academic publishing under the Open Access model.
Within this context, the Library of the University of Macedonia is conducting a
study aimed at mapping the views and experiences of the university’s
researchers regarding the advantages and challenges they encounter when
publishing in Open Access journals. The results of this research are expected to
significantly contribute to the development of informed policies and support
services tailored to the actual needs of the research community.
31o Πανελλήνιο Συνέδριο Ακαδημαϊκών Βιβλιοθηκών – 22-24/10/2025 Ιωάννινα
Το κείμενο της ανακοίνωσης (conference paper) είναι διαθέσιμο στο:
https://olympias.lib.uoi.gr/jspui/handle/123456789/39352
In addition, the presentation includes the findings of an analysis of the journals
in which researchers from the University of Macedonia published during the
period 2020–2024, compared against the Finnish journal evaluation system Jufo
(Julkaisufoorumi). The goal is to explore to what extent journals with a high
CiteScore in the Scopus database are also considered high-quality according to
Jufo's classification.
The analysis revealed that although most titles with a high CiteScore are ranked
in the higher Jufo tiers (levels 2 and 3), there are cases of journals with high
impact but low or no quality rating in Jufo (levels 1 and 0), indicating the need for
a multidimensional evaluation of publication.
Finally, the presentation highlights the need for institutional guidance and
support for researchers, the development of Open Access policies that ensure
the quality of scholarly communication, and the strengthening of the role of
libraries as hubs of support, evaluation, and awareness about Open Access
publishing
Beyond barriers: exploring foreign language learning experiences of students with diverse learning needs in four European countries
No full textWhile foreign language learning is increasingly recognized as crucial for educational and social inclusion, the experiences of students with diverse learning needs in foreign language classrooms remain understudied. This study investigated the relationship between Personal Engagement (PE) and Learning Attitudes (LA) among students with diverse learning needs in foreign language learning contexts across four European countries (Greece, Germany, Slovenia, and Poland). The study involved 95 students (aged 8-25) with various learning needs: visual impairment (n = 16), deafness/hard of hearing (n = 14), physical/motor impairment (n = 32), and learning difficulties (n = 33). Data were collected through interviews and standardized questionnaires examining both PE and LA, with findings analyzed using both qualitative and quantitative methods. Results revealed that LA scores consistently exceeded PE scores across all groups, with students with physical impairments showing the strongest correlation between engagement and attitudes (r = 0.674, p < 0.001), while students with visual impairments demonstrated high LA despite moderate engagement levels. Students with diverse learning needs maintain remarkably positive attitudes toward foreign language learning despite varying engagement levels, suggesting that educational barriers may be more related to access and delivery methods than to students’ willingness to learn. This emphasizes the need for tailored support strategies that can transform positive attitudes into fuller engagement across different types of learning needs.1
Perceptions of generative AI in education: Insights from undergraduate and master’s-level future teachers
No full textAs AI continues to transform educational practices, understanding its potential benefits, limitations, and ethical implications is critical. This study explores future educators‘ perceptions of generative ΑΙ in higher education. Through a mixed-methods approach, this research gathered insights from undergraduate and master‘s level students on their familiarity with, willingness to use, and concerns about AI-driven educational tools. Findings highlight a general positivity toward AI‘s utility in learning and teaching, with key advantages including personalised feedback, time efficiency, and academic support. Conversely, challenges related to reliability, academic integrity, and the potential for over-reliance reveal areas for further guidance and digital literacy enhancement. Post-intervention data suggests that hands-on experience with AI can reduce concerns and foster a balanced perspective on AI adoption in education. This study underscores the need for ethical frameworks and policy guidance to support the responsible integration of AI in educational contexts.928910
Investigating the Factors Influencing Teachers’ Intention to Use Chatbots in Primary Education in Greece
No full textThe usage of artificial intelligence (AI) in education is quickly growing, with chatbots gaining popularity as potential tools for supporting teaching and learning. This study looks into the elements that influence teachers’ willingness to use chatbots in their teaching techniques. Drawing on the Unified Theory of Acceptance and Use of Technology, the research focuses on performance expectancy (PE), effort expectancy (EE), and social influence (SI). In addition, it incorporates personal innovativeness (PI), perceived risk (PR), and personal experience with AI (PEAI) as supplementary variables. A quantitative approach was followed, using a questionnaire administered to 241 primary education school teachers in Greece. The proposed model was tested through partial least squares structural equation modeling. Findings reveal that PE is the strongest predictor of behavioral intention. PEAI positively affects both performance and EE and also has a direct impact on intention. EE and SI are also significant positive predictors, whereas PR negatively affects intention. Although PI does not directly influence intention, it contributes indirectly by enhancing perceived usefulness and ease of use and by lowering PR.712025010
Factors Affecting Internet Users’ Effectiveness to Detect Phishing
No full textPhishers target Internet users and send them phishing emails or create phishing websites to deceive victims and extract financial gains. This study investigated the factors that affect Internet users in phishing detection. The responses of 252 participants to an online survey were collected and analysed. Participants were mostly careful when a website requested urgent action, its content had errors, or an email message asked recipients to give personal information or click a suspicious link. They believed that education about phishing, email filtering software, and connecting directly to the official website, and not through a link, were the best practices. They also stated that they could learn how to easily detect phishing attempts and evaluate whether an email was spam, advertising, phishing, or a scam. It was also found that the creation of blacklists and whitelists did not provide sufficient protection against phishing, active warnings were more effective than passive warnings, and users were receiving phishing emails that escaped email filtering and antivirus software. Further statistical analysis of their responses revealed that there was a moderate positive correlation between each of the variables Self-Efficacy in Phishing Detection, Information Evaluation Skills, E-Banking Phishing Detection Skills with Phishing Detection Effectiveness. Finally, it was found that the Facilitating Conditions for Phishing Detection, E-Banking Phishing Detection Skills, and E-Mail Phishing Detection Skills influenced Phishing Detection Effectiveness.244255002
Spectral Unmixing for Bare Soil Detection: A Step Toward Soil Organic Carbon Estimation
No full textThis study presents a benchmark for the detection
of bare soil from EnMAP hyperspectral imagery, addressing the challenge of mixed pixels, a known limitation in satellitebased soil property estimation. A reference bare soil mask was constructed using a histogram separation method applied to
spectral indices of a multitemporal Sentinel-2 composite. Using this higher resolution reference, we evaluate two classification approaches on EnMAP data: spectral index thresholding and linear spectral unmixing. The latter achieves higher classification
performance, indicating improved reliability under heterogeneous surface conditions. The benchmark dataset created for this study is publicly available.142025 International Conference on Machine Intelligence for GeoAnalytics and Remote Sensing (MIGARS
Transfer learning for software vulnerability prediction using Transformer models
No full textRecently software security community has exploited text mining and deep learning methods to identify vulnerabilities. To this end, the progress in the field of Natural Language Processing (NLP) has opened a new direction in constructing Vulnerability Prediction (VP) models by employing Transformer-based pre-trained models. This study investigates the capacity of Generative Pre-trained Transformer (GPT), and Bidirectional Encoder Representations from Transformers (BERT) to enhance the VP process by capturing semantic and syntactic information in the source code. Specifically, we examine different ways of using CodeGPT and CodeBERT to build VP models to maximize the benefit of their use for the downstream task of VP. To enhance the performance of the models we explore fine-tuning, word embedding, and sentence embedding extraction methods. We also compare VP models based on Transformers trained on code from scratch or after natural language pre-training. Furthermore, we compare these architectures to state-of-the-art text mining and graph-based approaches. The results showcase that training a separate deep learning predictor with pre-trained word embeddings is a more efficient approach in VP than either fine-tuning or extracting sentence-level features. The findings also highlight the importance of context-aware embeddings in the models’ attempt to identify vulnerable patterns in the source code.22711244