Browsing by Author "Ozarslan, Yasin"

Now showing 1 - 8 of 8

Citation - WoS: 1
Citation - Scopus: 1
A Feature Selection Application Using Particle Swarm Optimization for Learning Concept Detection
(SPRINGER-VERLAG BERLIN, 2017) Korhan Gunel; Kazim Erdogdu; Refet Polat; Yasin Ozarslan; Polat, Refet; Erdogdu, Kazim; Ozarslan, Yasin; Gunel, Korhan; A Rocha; AM Correia; H Adeli; LP Reis; S Costanzo
Recent developments of computational intelligence on educational technology yield concept map mining as a new research area. Concept map mining covers the extraction of learning concepts specifying relations among them and generating a concept map from educational contents. In this study we focused on determining the features that characterize a learning concept extracted from an educational text as raw data. The first three features are detected by using a hybrid system of Multi Layer Perceptron (MLP) and Particle Swarm Optimization (PSO) and the performance of the applied method is gauged in the viewpoint of a typical classification problem.
Citation - WoS: 8
Citation - Scopus: 10
An empirical study on evolutionary feature selection in intelligent tutors for learning concept detection
(WILEY, 2019) Korhan Gunel; Kazim Erdogdu; Refet Polat; Yasin Ozarslan; Polat, Refet; Erdogdu, Kazim; Ozarslan, Yasin; Gunel, Korhan
Concept map mining (CMM) has emerged as a new research area with recent developments in computational intelligence in educational technology. CMM includes the following steps: extracting the learning concepts from educational content specifying relations among them and generating a concept map as a result. The purpose of this study was to develop a mechanism using data mining technique to determine the features that characterize a learning concept extracted automatically from a single educational text. The 3 major features that distinguish the real learning concepts from other sequences of strings are detected by using a hybrid system of a feed-forward neural network and some evolutionary algorithms. Ant colony optimization and genetic algorithm and particle swarm optimization are used as a binary feature selection method. In addition the aforementioned methods are hybridized to get better accuracy and precision. The performance comparisons with two different state-of-the-art algorithms have been made from the viewpoint of a typical classification problem.
Automatic Short-Answer Grading in Sustainability Education: AI-Human Agreement
(Wiley, 2026) Emirtekin, Emrah; Ozarslan, Yasin
Background Sustainability education emphasises critical thinking and interdisciplinary understanding, making the assessment of students' learning outcomes complex. While Large Language Models (LLMs) have shown promise in educational assessment, their reliability in domains requiring contextual reasoning-such as sustainability-remains unclear. Objectives This study aims to evaluate the agreement between human raters and several LLMs (GPT-4o, Gemini 2.0 Flash, DeepSeek V3, LLaMA 3.3) in assessing short-answer responses from a university-level Sustainability course. It also investigates how this agreement varies across cognitive skill levels. Methods A total of 232 short-answer responses were evaluated using a rubric aligned with Bloom's Revised Taxonomy. Consensus scores from human raters were compared to LLM-generated scores using multiple statistical measures, including Quadratic Weighted Kappa (QWK), Intraclass Correlation Coefficient (ICC), Pearson correlation, and distributional overlap. Results Moderate agreement was found between LLMs and human raters in total scores (QWK: 0.585-0.640; r: 0.660-0.668; eta: 0.681-0.803). Inter-rater reliability among humans was good to excellent (ICC: 0.667-0.800). Criterion-level agreement declined as cognitive complexity increased, with notably low agreement on evaluating higher-order skills. Conclusions Overall, LLM-human agreement was moderate on total scores but declined at higher cognitive levels, indicating that LLMs are suitable for basic comprehension checks while human oversight remains necessary for complex reasoning.
Citation - Scopus: 2
Factors Influencing the Learner's Cognitive Engagement in a Language MOOC: Feature Selection Approach
(Institute of Electrical and Electronics Engineers Inc., 2023) Murat Kılınç; Orkun Teke; Ozlem Ozan; Yasin Ozarslan; Kilinc, Murat; Teke, Orkun; Ozan, Ozlem; Ozarslan, Yasin
This study aims to predict the cognitive engagement rate in a Language MOOC (Massive Open Online Course) based on the features extracted from learners' engagement behaviors within the content and activities. The features were extracted from the data of the Language MOOC 'Türkçe Öǧreniyorum (I learn Turkish)' which aims to provide self-paced learning materials for those interested in developing their skills in Turkish as a foreign language. After the data preprocessing processes were carried out with the data set obtained for cognitive engagement classification feature selection processes were performed using filtering and wrapper methods. Afterward the machine learning model trained using the Logistic Regression (LR) algorithm performed the classification with 94% accuracy. The model evaluation metrics also support the classification result obtained. Based on the extracted features and the classification results obtained the model will be able to capture learners' interaction behaviors with the content and activities in a Language MOOC and detect changes in learner behavior over time. Prediction accuracy is essential to offer dynamic content and activities in a Language MOOC for adjusting the individual needs of each learner providing personalized learning experiences that are tailored to their skills knowledge and preferences. © 2023 Elsevier B.V. All rights reserved.
Citation - WoS: 6
Citation - Scopus: 8
Gen Z travel intentions and museum visits in the metaverse: case of Egypt- Scotland- and Turkey
(ROUTLEDGE JOURNALS TAYLOR & FRANCIS LTD, 2025) Murat Nazli; Cagri Bulut; Yasin Ozarslan; Nazli, Murat; Bulut, Cagri; Ozarslan, Yasin
This early-stage study explores Gen Z's travel intentions and virtual museum visits in the metaverse by experiencing Egypt Scotland and Turkey. The subjective perception and judgment of Gen Z users are investigated in the immersive environment. Semi-structured interviews and laboratory sessions took place with 20 volunteers in three studies. Study 1 and Study 2 reveal different views of Gen Z concerning their travel intentions as two open-air experiences in the metaverse. Study 3 shows Gen Z perspectives of a virtual museum visit as an indoor experience. Through this study we could form a link between the metaverse and Gen Z travel intentions using two travel applications for Egypt and Scotland and a museum application for Turkey which has rarely been discussed before. The study also guides an experience interaction interface design for VR 360 video and future-related tourism products and services for Gen Z users. The study illuminates the travel and metaverse relationship for future studies for practitioners and researchers.
Supporting SDG-Oriented Knowledge Construction and Idea Diffusion in Online Higher Education
(MDPI, 2026) Ozarslan, Yasin; Ozan, Ozlem
This study investigates how online discussion forums in an undergraduate Social Responsibility course support students' SDG-oriented idea generation and collaborative knowledge construction. It also examines how participation roles, behavioral intensity, interaction-network influence, and goal-aligned discourse shape idea visibility and discussion. Using a mixed-methods learning analytics design, we analyzed forum logs and message texts across five SDG-linked themes (SDGs 6, 7, 12, 14, 15) by classifying contributor types, computing a Behavioral Participation Index (BPI), constructing a directed reply network and estimating PageRank centrality, extracting solution proposals, scoring semantic goal alignment, modelling weekly temporal dynamics, and fitting multivariate regressions predicting visibility (reads) and engagement (replies) while controlling for theme, message level, time, PageRank, and BPI. Results show role-differentiated participation (N = 514), meaningful cross-theme solution proposals that varied across academic groups, and peak-driven weekly activity. PageRank centrality emerged as the strongest and most consistent predictor of both visibility and engagement, whereas goal alignment showed weaker direct effects after controls, suggesting that SDG-aligned ideas do not necessarily diffuse without structural embeddedness. Among highly goal-aligned posts, specific communicative features differentiated which proposals attracted attention and interaction. These findings suggest that SDG forum design benefits from structured interaction pathways and scaffolded discourse strategies to support equitable diffusion and productive sustainability dialogue. The study does not evaluate the normative quality of sustainability positions but examines how interaction structures and discourse features shape the visibility and diffusion of student-generated ideas.
Citation - WoS: 1
Citation - Scopus: 1
TOWARDS AN ADAPTIVE LANGUAGE MOOC: EXAMINING DIFFERENCES OF LANGUAGE ERROR PATTERNS ACROSS CULTURAL DOMAINS
(Anadolu Universitesi, 2025) Ozlem Ozan; Yasin Ozarslan; Sevgi Calisir Zenci; Calisir Zenci, Sevgi; Ozan, Ozlem; Ozarslan, Yasin
This study analyzed linguistic errors as part of the Differentiated Distance Education of Turkish as a Foreign Language Project which pursues the development of an adaptive MOOC for Turkish as a second language. Therefore the Turkish CEFR (Common European Framework of Reference for Languages) A1-level writing exam papers of 177 learners were analyzed. Linguistic error analysis techniques were used. A Chi-square test of independence a Kruskal-Wallis H test and a Mann-Whitney U test were conducted to examine the data. The results show a relationship between error frequency and learner group (Arabic–Farsi Turkic Balkan and Other). Similarly the error density varied as a function of the learner group. There is also a relationship between error frequency and the language family of the learner’s mother language. On the other hand there is no significant difference in error density by language family. The number of languages the learner knows has no significant effect on error frequency and density. The findings suggest that there are gender-based differences in error density among learners but that these differences are not reflected in the frequency of errors. The topics for differentiation were identified based on the error distribution of learner groups. The topic that requires the most differentiation is noun phrases. The learner groups that need the most differentiation are the Arabic and Farsi Nations while the Turkic Nations require the least differentiation. © 2025 Elsevier B.V. All rights reserved.
Citation - WoS: 50
Citation - Scopus: 63
Video lecture watching behaviors of learners in online courses
(Routledge info@tandf.co.uk, 2016) Ozlem Ozan; Yasin Ozarslan; Ozan, Ozlem; Ozarslan, Yasin
This paper examines learners’ behaviors while watching online video lectures to understand learner preferences. 2927 students’ 18144 video events across 13 courses on Sakai CLE LMS which were integrated with Kaltura Video Platform and Google Analytics were analyzed. For the analysis of the quantitative data one-way ANOVA Chi-square test of independence and descriptive statistics were utilized. The main results revealed that there was a tendency toward watching interview-style video lectures completely. In addition the percentage rate of Watching Completely behavior was higher in shorter videos and a tendency toward watching long video lectures by seeking was found. According to our results watching patterns was also affected by lecturers’ characteristics. Watching Completely rate of female lecturers was significantly different than those of male lectures in favor of females as well as watching in FullScreen mode. Furthermore learners who watched online video lectures completely had higher scores on the final exam than others. Our analysis could help those who plan to optimize online video lectures in e-learning programs. © 2016 Elsevier B.V. All rights reserved.