Scopus İndeksli Yayınlar Koleksiyonu / Scopus Indexed Publications Collection

Permanent URI for this collectionhttps://hdl.handle.net/20.500.11779/1926

Browse

Now showing 1 - 20 of 715

The Role of Artificial Intelligence in Sustainable Food Retailing: A Pathway to Reducing Waste and Enhancing Efficiency
(Springer Science and Business Media B.V., 2026) Köse, Ş.G.; Kazançoǧlu, I.; Kazancoglu, Y.
Food waste arises in the supply chain’s distribution, marketing, and consumption phases. Retailers face problems related to food waste since it increases operating costs, causing savings and loss of profitability. Due to food waste’s environmental, economic, and social impacts, food retailers increasingly incorporating artificial intelligence-based applications into their operations to increase efficiency and ensure sustainability. Demand forecasting, inventory management and optimization, dynamic pricing, supply chain, and order optimization are some of the main artificial intelligence applications utilized to address food waste. In line with this purpose, the paper aims to highlight the importance of artificial intelligence for reducing food waste in retailing. This study focuses on how food retailers increasingly leverage artificial intelligence to reduce food waste and improve sustainability across their operations. This research was conducted through case study and these cases provide applications of artificial intelligence to the food retail industry. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2026.
Evaluating Large Language Models in Data Generation for Low-Resource Scenarios: A Case Study on Question Answering
(International Speech Communication Association, 2025) Arisoy, E.; Menevşe, M.U.; Manav, Y.; Özgür, A.
Large Language Models (LLMs) are powerful tools for generating synthetic data, offering a promising solution to data scarcity in low-resource scenarios. This study evaluates the effectiveness of LLMs in generating question-answer pairs to enhance the performance of question answering (QA) models trained with limited annotated data. While synthetic data generation has been widely explored for text-based QA, its impact on spoken QA remains underexplored. We specifically investigate the role of LLM-generated data in improving spoken QA models, showing performance gains across both text-based and spoken QA tasks. Experimental results on subsets of the SQuAD, Spoken SQuAD, and a Turkish spoken QA dataset demonstrate significant relative F1 score improvements of 7.8%, 7.0%, and 2.7%, respectively, over models trained solely on restricted human-annotated data. Furthermore, our findings highlight the robustness of LLM-generated data in spoken QA settings, even in the presence of noise. © 2025 International Speech Communication Association. All rights reserved.
Turkey's EU Membership Process in the Aftermath of the Gezi Protests
(Taylor and Francis, 2025) Saatçioǧlu, B.
Methane Emissions Forecasting Using Hybrid Quantum-Classical Deep Learning Models: Case Study of North Africa
(Springer, 2025) Belkadi, Widad Hassina; Drias, Yassine; Drias, Habiba; Ferkous, Sarah; Khemissi, Maroua
This study explores climate change by predicting methane emissions in North Africa using classical and quantum deep learning methods. Using data from Sentinel-5P, we developed hybrid quantum-classical models, such as quantum long short-term memory (QLSTM) and quantum-gated recurrent unit networks (QGRUs), along with a novel hybrid architecture combining quantum convolutional neural networks (QCNNs) with LSTM and GRU, namely QCNN-LSTM and QCNN-GRU. The results show that these quantum models, especially the proposed hybrid architectures, outperform classical models by approximately seven percent in root-mean-squared error with fewer training epochs. These findings highlight the potential of quantum methodologies for enhancing environmental monitoring accuracy. Future research will aim to refine model performance, incorporate explainable AI techniques, and expand to forecasting other greenhouse gases, contributing to climate change mitigation efforts.
Big-5 Personality Traits as Predictors of Allostatic Load in Latino Americans: A Longitudinal Study
(Oxford Univ Press Inc, 2025) Sevi, Baris; Supiyev, Adil; Gutierrez, Angela; Graham, Eileen K.; Mroczek, Daniel K.; Muniz-Terrera, Graciela
Objectives Allostatic load (AL) refers to the measure of cumulative wear and tear resulting from chronic stress and life events. AL presents adverse consequences for a diverse range of health conditions, and Latino populations show a high risk for elevated AL. This study aimed to test the Big-5 personality traits as possible predictors of AL in Latinos.Methods Using data from the Health and Retirement Study, we examined the Big-5 and AL connection through three time points in 8 years (Time 1 = 2006/2008; Time 2 = 2010/2012; Time 3 = 2014/2016). Only self-identified Latinos were included in the analysis sample (N = 319). Big-5 and demographics were obtained at baseline, and AL scores were computed for each time point.Results First, separate longitudinal linear mixed-effect models examined the effects of each Big-5 personality trait on AL change over time, then a fully adjusted longitudinal linear mixed-effect model was tested entering the Big-5 personality traits simultaneously. All models controlled for sociodemographic factors. Conscientiousness emerged as the only consistent significant predictor, for the separate and the simultaneous models. In baseline associations, higher conscientiousness was associated with lower AL. For predicting change in AL over time, none of the personality traits had significant associations in any of the models.Discussion The findings bolster prior evidence that conscientious can be a protective factor against elevated AL. Conscientiousness is a possible protective factor and improving related traits can be a path to achieve better health in Latino Americans.
Moral Framing Effects on Environmental Attitudes: A Conceptual Replication and Extension of Feinberg and Willer (2013)
(Academic Press Ltd- Elsevier Science Ltd, 2025) Cavdar, Dilara; Tepe, Beyza; Saribay, S. Adil; Yilmaz, Onurcan
This study investigates the relationship between moral framing, political orientation, and pro-environmental attitudes, replicating and extending Feinberg and Willer (2013) in a non-Western context. Using a Turkish-speaking sample (N = 699), we examined the effectiveness of care and sanctity-framed messages and the moderating role of actively open-minded thinking (AOT). Our findings partially replicated the original study. Sanctity framing increased pro-environmental attitudes among conservatives, while care framing had no significant effect. Political conservatism was negatively associated with pro-environmental attitudes, confirming prior findings. Exploratory analyses revealed that AOT moderated the effects of sanctity framing on environmental attitudes, with individuals low or moderate in AOT being more responsive. Both care and sanctity frames increased environmental donation, addressing the intention-behavior gap. However, cultural nuances, such as the collectivist orientation of the sample, may have influenced the care frame's ineffectiveness. The study highlights the importance of cultural context in moral framing research and underscores the need for context-specific climate communication strategies.
Enhanced Knowledge Transfer in Historical Site Interpretation: A Gamified Approach for Yedikule Fortress/Türkiye
(Springer Science and Business Media Deutschland GmbH, 2025) Özer, D.G.; Gül, L.F.; Coskun, E.; Kilimci, E.S.Y.; Gül, M.
This study explores the integration of wearable and portable devices in enhancing the interpretation and communication of Yedikule Fortress, a 1600-year-old historical site on the outskirts of the Istanbul’s Historic Peninsula. This interpretation involves users taking of the site tour with augmented reality/mixed reality-enhanced using audio accompaniment and gamification. The designed route provides tailored information, enhanced by a digital layer and a storyline. This study examines the quality of the information, visitor route design, and the integration of game elements. The multi-modal guidance system fosters interaction between historical narratives and spatial context, with participants reporting increased interest and improved recall of historical details. Building on the refinement of the process through expert input and participant feedback, this study also highlights limitations related to device affordances that influenced user behavior and interaction. Future research should explore the adaptability of this framework to other heritage sites, ensuring its educational and experiential relevance. © 2025 Elsevier B.V., All rights reserved.
Quantum FP-Growth Algorithm Using GPU Simulation-Application to Digital Soil Mapping
(Elsevier, 2026) Belkadi, Widad Hassina; Drias, Yassine; Drias, Habiba
This study introduces a novel quantum version of the FP-growth algorithm for frequent itemset mining, leveraging the combined strengths of classical FP-growth and quantum machine learning. Key contributions include the theoretical and practical framework for Quantum FP-growth, along with a comprehensive analysis of its time and space complexity. We implemented Quantum FP-growth using IBM Qiskit and conducted a comparative evaluation of various quantum amplitude estimation (QAE) methods, including Canonical QAE, Faster QAE, Maximum Likelihood QAE, and Iterative QAE for support estimation. Our findings reveal that Iterative QAE surpasses the other methods in both accuracy and speed. Additionally, we explored the advantages of GPU simulation with IBM Qiskit and NVIDIA cuQuantum. Notably, this research marks the first application of a quantum frequent itemset mining algorithm to a real-world dataset in Digital Soil Mapping (DSM), pioneering the use of quantum technologies in soil science. This study underscores the potential of quantum computing to revolutionize data mining and promote sustainable soil management practices.
Exploring Generative AI and Unsupervised Learning for Digital Soil Classification: A Case Study of Algeria
(Institute of Electrical and Electronics Engineers Inc., 2025) Belkadi, W.H.; Drias, Y.; Drias, H.; Bouchelkia, H.; Hamdous, S.
Accurate soil type mapping is vital for sustainable agriculture and land management. Yet, Algeria remains under-represented in global soil databases. To address this, we propose a pipeline combining Generative AI for data extraction with unsupervised learning for soil classification. After harmonizing Algerian soil data, we evaluate four clustering algorithms - K-means, DBSCAN, HDBSCAN, and Self-Organizing Maps - under various preprocessing settings. Internal and external metrics guide model selection. K-means and DBSCAN produced the most coherent clusters, while SOM best aligned with FAO soil types. RuleFit was then used to extract interpretable rules defining each cluster. This work highlights the potential of AI-based, interpretable clustering for digital soil mapping in data-scarce regions like Algeria. © 2025 Elsevier B.V., All rights reserved.
Burdens of Masculinity Among Heterosexual, Gay, and Bisexual Men in Turkey: More Masculine, More Conflicted, Less Satisfied
(Springer, 2025) Toplu-Demirtas, Ezgi; Oztemur, Gizem; Keskin, Berat; Fincham, Frank D.
Although bivariate associations among masculinity ideology, gender role conflict, and life satisfaction have been documented in Western countries, they have received limited attention in Turkey. Moreover, the majority of peer-reviewed research on masculinity has focused on heterosexual men's experiences. The current study, therefore, explored the relationship between masculinity ideology and life satisfaction in Turkish men with gender role conflict as a mediator and sexual orientation (heterosexual men vs. gay or bisexual men) as a moderator variable. Data were collected online from 195 men (128 heterosexual, 53 gay, and 14 bisexual) between the ages of 18 and 42 (M = 25.39, SD = 3.53) using the Life Satisfaction Scale, Masculinity Ideology Scale, and Gender Role Conflict Scale. The moderated-mediation analysis revealed that masculinity ideology and life satisfaction were significantly associated via the mediator of gender role conflict. Both heterosexual and gay or bisexual men who adhered more to masculine ideology experienced greater gender role conflict and thus felt less satisfaction with life. After discussing the results and their limitations, recommendations for further research and practice are offered. We conclude that addressing gender role conflict in clinical work may be a profitable approach to increasing men's life satisfaction.
Heterogeneous Impact of Innovation on Economic Development: Evidence from EU Regions
(Elsevier Sci Ltd, 2026) Pinar, Mehmet; Karahasan, Burhan Can
This paper investigates the heterogeneous impact of innovation on economic development across European Union (EU) regions, with a focus on regional competitiveness driven by innovation-based capabilities. While innovation is a key driver of economic growth, its effects are not uniformly distributed. Using the Multiscale Geographically Weighted Regression models, the study examines how different dimensions of innovation (technological readiness, business sophistication, and overall innovation capacity) affect regional GDP per capita. The results show that regions with higher innovation-based competitiveness generally achieve higher income levels. However, the impact of innovation is spatially uneven. While core EU regions (particularly, in Northern and Western Europe) benefit more strongly from innovation, peripheral regions (in Southern and Eastern Europe) often experience weaker and in some cases even negative, effects. These results highlight the importance of accounting for spatial variation when designing innovation and cohesion policies. The paper calls for tailored, place-based strategies to address regional disparities in innovation-driven development and suggests that current EU policies should be adjusted to better support lagging regions.
Ruling Through Exception: Lawfare, Securitised Warfare and the Intermestic Logic of Authoritarianism
(Routledge Journals, Taylor & Francis Ltd, 2025) Çağlar, Barış
This article develops an original interdisciplinary framework for analysing authoritarian regimes. It coins, for the first time, the concept of securitised warfare, theorising its conceptual foundations, and it also originates and develops an original theoretical framework synthesising securitisation, authoritarianism and structuration. Securitised warfare - defined here as the outward intermestic manifestation of lawfare - is shown to be mutually constitutive with lawfare, the strategic misuse of the legal system for political gain, with both reinforcing the consolidation of authoritarian rule. Focusing on Turkey (2015-2025), the article illustrates how the regime employed legal repression as a political instrument, particularly in the cases of Selahattin Demirta & scedil; and Ekrem & Idot;mamo & gbreve;lu. Simultaneously, the suppression of Kurdish groups in Syria exemplifies securitised dynamics shaped in conjunction with domestic politics. Using Lijphart's hypothesis-generation method and within-case process tracing, the study demonstrates how lawfare and securitised warfare function both as Schmittian exceptions and as routinised Giddensian institutional practices. The framework conceptualises the historical transition from national security state to neoliberal security state, culminating in the consolidation of an autocratic regime whose logic exceeds conventional regime security. This transformation is theorised through securitised warfare - explaining how domestic and foreign policy are increasingly governed by a unified logic of authoritarian control.
The TRAIN Framework for AI-Assisted Development of Novice Accounting Instructors
(Springer Science and Business Media Deutschland GmbH, 2025) Son-Turan, Semen
Novice accounting instructors in higher education, and particularly in specialized fields like accounting, face significant challenges, including the need to frequently update curricula, balance theoretical and practical instruction, master new technologies, and integrate sustainability principles into their teaching. These demands are compounded by limited access to professional development resources and growing pressures for educational sustainability. This study evaluates the potential of AI chatbots to address these challenges by enhancing the capacity of junior instructors to create and deliver effective, sustainability-aligned course materials. It introduces the TRAIN framework (Technology-enhanced Responsible AI Integration for Novice Instructors), a structured model designed to align AI capabilities with Sustainable Development Goals (SDGs), particularly SDG 4 (Quality Education) and SDG 4.c (Teacher Training). Through a systematic literature review (2019–2024), the research identifies key themes, including chatbot integration in education, AI’s capabilities and limitations, and its broader implications for accounting education. The findings demonstrate that AI chatbots can streamline content creation, provide personalized instructional support, and foster sustainable teaching practices. By bridging critical gaps in AI literacy and sustainability integration, the TRAIN framework offers a roadmap for empowering educators to meet evolving educational demands while advancing environmentally responsible teaching methodologies. © 2025 Elsevier B.V., All rights reserved.
Makine Öğrenimi ve Çok Boyutlu Anket Verileri Kullanılarak Öğrenci Başarısının Tahmini: Eğitim Programı Üzerine Bir Uygulama
(Institute of Electrical and Electronics Engineers Inc., 2025) Behsi, Zeynep; Dereli, Serhan; Çakar, Tuna ; Patel, Jay Nimish; Cicek, Gultekin; Drias, Yassine
This study develops a machine learning model integrating survey data and performance metrics to predict student success in the UpSchool education program. Students' personality traits assessed by DISC analysis, financial management, social, and emotional skills were clustered into "Successful,""Unsuccessful,"and "Moderately Successful"groups using K-means clustering. The SMOTE technique addressed data imbalance issues, and algorithms such as Logistic Regression, Random Forest, LightGBM, and AdaBoost were tested. After hyperparameter optimization, AdaBoost and LightGBM achieved the highest predictive performance. Results demonstrated the effectiveness of machine learning models in forecasting student success in educational programs. Future studies are recommended to enhance model performance through expanded datasets and to validate the model's applicability across diverse educational contexts. © 2025 Elsevier B.V., All rights reserved.
İnternet Trafik Hızının Tahmininde Derin Öğrenme ve Ağaç Tabanlı Modellerin Karşılaştırılması
(Institute of Electrical and Electronics Engineers Inc., 2025) Filiz, Gozde; Altıntaş, Suat; Yıldız, Ayşenur; Kara, Erkan; Drias, Yassine; Çakar, Tuna
This study addresses the prediction of internet traffic speed using time-dependent data from an internet service provider through different modeling approaches. On an anonymized dataset, the performance of the moving average method, various deep learning models (N-BEATS, N-HITS, TimesNet, TSMixer, LSTM), and the XGBoost regression model enhanced with feature engineering was compared. Time series cross-validation and random hyperparameter search were used for model training. According to the results, the XGBoost model achieved the highest accuracy with 98.7% explained variance (R2), while among the deep learning models, N-BEATS and N-HITS achieved the best performance with R2 values around 90%. The findings indicate that tree-based methods supported by carefully selected features can offer higher accuracy and computational efficiency compared to complex deep learning models in internet traffic forecasting. © 2025 Elsevier B.V., All rights reserved.
Financial Inputs Prediction with Machine Learning and Covariance Matrix Applications
(Institute of Electrical and Electronics Engineers Inc., 2025) Benli, Harun; Gunes, Peri; Ulkgun, Mert; Cakar, Tuna
Reliable estimation of the time-varying covariance matrix of asset returns is indispensable for portfolio construction, risk budgeting, and automated advisory services. Conventional estimators-rolling-window sample covariances, EWMA filters, and GARCH families-remain anchored to historical prices and therefore adapt slowly when market conditions pivot. To overcome this latency, we propose an end-to-end, machine-learning-driven framework that forecasts future covariances directly from high-frequency equity data, largely decoupling risk estimation from past observations. Our pipeline ingests heterogeneous stock feeds through a real-time API, applies error-minimising imputation (forward/backward fill, spline, VAR, wavelet, and co-kriging), and standardises returns via empirically selected scaling schemes. The processed features are then fed to a model zoo comprising linear and penalised regressions, tree ensembles (Random Forest, XGBoost, LightGBM, CatBoost), and kernel-based Support Vector Regression. Weekly walk-forward evaluation on a universe of Turkish insurance equities shows that LightGBM and SVR cut out-of-sample covariance prediction error by up to 35 % versus classical benchmarks. We embed the predicted matrices into five allocation engines-Markowitz mean-variance, Black-Litterman, minimum-variance, Risk Parity, and CVaR optimisation-demonstrating that Risk Parity delivers the most consistent variance reduction across 15 stock pairs, while Black-Litterman excels for idiosyncratic combinations such as ANSGR-AKGRT. A granular analysis reveals that day-to-day sign changes in returns create structural breaks that generic regressors miss; augmenting the architecture with a volatility-state classifier and regime-specific learners markedly sharpens turning-point detection. Beyond statistical gains, the framework is production-ready: it is fully implemented in Python, runs on cloud notebooks, and plugs into robo-advisory dashboards. The study thus bridges academic advances in covariance prediction with operational portfolio management, paving the way for broader cross-sector deployment and future research on deep sequential models, transaction-cost awareness, and multi-asset scalability. © 2025 Elsevier B.V., All rights reserved.
Yapay Öğrenme Tabanlı Mikrofaktoring Skorlama Modeli ve Kredi Risk Yönetim Sistemi Geliştirilmesi
(Institute of Electrical and Electronics Engineers Inc., 2025) Sayar, Alperen; Ates, Yigit; Ertugrul, Seyit; Turan, Elif Naz; Drias, Yassine; Çakar, Tuna
Credit scoring systems are critical tools used by factoring institutions to assess the credit risks of SME businesses seeking microloans. This study presents a comprehensive predictive modeling framework that achieves 82.67% ROC-AUC with 65.34% Gini score on test data, demonstrating robust discriminative capability despite significant class imbalance. Our ensemble approach outperforms individual boosting models by leveraging their complementary strengths in payment behavior analysis and fraud detection. The raw data was cleaned, transformed, and optimized using the Polars library, with specialized features for detecting fraud patterns and time-based risk indicators. When implementing a score threshold of 950, our model significantly improves the detection of non-performing loans (NPL) compared to traditional rule-based approaches by reducing the net deficit from 6.59% to 2.62%. When applied to previously rejected applications, the model projects a potential 762.57% increase in transaction count and 747.05% growth in transaction volume. © 2025 Elsevier B.V., All rights reserved.
Ensemble-Based Stock Prediction for Retail - XGBoost and LightGBM with Rolling Window Training
(Institute of Electrical and Electronics Engineers Inc., 2025) Patel, Jay Nimish; Kizilay, Ayse; Şahin, Zeynep; Sercan, Busra; Toprak, Samet; Çakar, Tuna
Stock prediction in retail settings is a critical challenge that impacts numerous businesses globally, that require precise and timely forecasts to optimize inventory management and enhance customer satisfaction. State-of-the-art approaches for accurate stock prediction leverage machine learning (ML) models, which require large amounts of historical sales data for effective training. Such detailed datasets are often hard to obtain, limiting the performance and scalability of these approaches. In this paper, we propose various strategies to tackle this limitation. Initially, we adopt a transfer-learning approach, utilizing pre-trained models like XGBoost and LightGBM, which are fine-tuned for stock prediction in retail environments. To further boost model performance, we incorporate an ensemble method that combines predictions from both models to improve accuracy and manage outliers. Experiments conducted on an extremely large dataset, comprising millions of retail transactions, highlight the presence of significant outliers. Our models, augmented with ensemble strategies, significantly outperform traditional models in handling these complexities and improving prediction accuracy. © 2025 Elsevier B.V., All rights reserved.
Frekans Seçici Yüzeyler ile Mikroşerit Yama Antenin Radar Kesit Alanının Azaltılması
(Institute of Electrical and Electronics Engineers Inc., 2025) Metin, Ahmet Resul; Aydin, Irem; Bilgin, Egemen
In this study, the radar cross section (RCS) of a microstrip patch antenna was reduced via frequency selective surfaces (FSS). Since microstrip antennas are frequently used in military platforms, antenna design with low radar cross section will contribute to the general efforts towards reducing RCA. First, a microstrip antenna with a center frequency of 5 GHz was designed and a FSS that acts as a reflector at this frequency was developed. Here, the expectation from the FSS is to act as a reflector at the center frequency and to have a transmissive behavior at other frequencies, contributing to the reduction of the radar cross section. By integrating the designed FSS structure into the ground plane of the antenna, RCA reduction was achieved in different bands without shifting the center frequency. According to the simulations, the designed antenna has a 10 dB lower radar cross section in the X band. © 2025 Elsevier B.V., All rights reserved.
Suggestions in Digital Discourse: The Case of MOOC Reviews
(Elsevier B.V., 2025) Ciftci, Hatime
This study examines the speech act of suggestions in digital discourse through linguistic and functional approaches and explores how suggestions are performed along with co-occurring discourse-pragmatic particles, supporting moves, and aspects in their propositional content. More specifically, this paper presents findings regarding the speech act of suggestions in MOOC reviews as a recent and emerging genre of digital discourse. Embracing a discourse analytic perspective, this study indicates how suggestions are situated within the context they are used, and their multi-functionality is evidently relevant to the linguistic choices and supporting moves by MOOC learners, going beyond the utterance level meaning. Additionally, suggestion head acts involve certain aspects of online courses or their experience where learners often express their expectations or opinions for improvement. Overall, this study contributes to speech act research in digital discourse and provides insights into the use of suggestions in the discourse of MOOC reviews. © 2025 Elsevier B.V., All rights reserved.

Browse

Recent Submissions