5916 research outputs found
Sort by
Developing a robust question and answer system for the Turkish language utilizing deep learning techniques
No full textResearch on open-ended question and answer systems faces several complex challenges within the domain of natural language processing, one of which is limited data availability. Although numerous studies have been conducted in widely spoken languages such as English, there is a notable scarcity of research in languages such as Turkish. In our preliminary investigation, we proposed a solution to address the data shortage issue in Turkish by translating the English SQuAD dataset into Turkish. Another challenge is the success of deep learning models that use large language models. We developed various baseline models using deep learning techniques on this newly created dataset and performed analyses from multiple perspectives. Deep learning models and large language models often present an architectural enigma for many researchers, so analyzing both the questions and the corresponding answer-bearing data is of utmost importance. We have shown that the structuralization and enrichment of the data contribute significantly to the success of the model. In our research, we devised an architecture that incorporates a structural transformation of data before use in model training. This approach enabled us to enhance the learning capacity of the system without altering the underlying closed-box architecture of the large language models and deep learning systems employed.Publisher versio
Real-time ground fault detection in ECUs with high-side switches driving inductive loads
No full textAutomotive electronic control units (ECUs) often control inductive loads such as solenoids, electromagnetic retarders, relays, or electromagnetic actuators via high-side MOSFET switches powered by charge pumps. In typical 12V or 24V systems, these configurations are widely used in applications such as retarder control, valve actuation, and power relay switching. While conventional protection mechanisms address short-circuit and open-load conditions, a critical yet often overlooked fault arises when the ECU’s ground is lost, causing excessive voltage differentials that can overstress protection elements like TVS diodes and damage voltage regulators or microcontroller peripherals. This fault may also disrupt CAN communication and impair ECU functionality. This paper proposes a practical and cost-effective real-time diagnostic method, implemented within an automotive embedded system, using a simple voltage monitoring network and microcontroller-based signal processing. Compared to existing methods, the proposed scheme offers a low-cost and easily integrable solution that enables real-time ground-loss detection, ensuring faster fault response and improved functional safety
Digital accountability through e-participation: The moderating role of the digital divide
No full textThe authors show that equitable access to digital technologies enhances the effectiveness of e-participation, enabling citizens to monitor, evaluate, and influence public sector decision-making more meaningfully. The findings presented in this article have important implications for policy-makers-infrastructure expansion alone cannot achieve digital accountability. Instead, digital inclusion and civic engagement must be addressed through investments in broadband access, digital literacy, and participatory platforms. Bridging the digital divide ensures that accountability mechanisms are accessible to all, thus promoting transparency, responsiveness, and trust in government. By connecting digital transformation to participatory governance, this article supports efforts toward sustainable and inclusive public administration, offering guidance for building citizen-centreed digital governance frameworks.
The authors investigated the relationship between e-participation and digital accountability, with a particular focus on how the digital divide moderated this relationship across the EU 27 between 2018 and 2022. Using panel data Feasible Generalized Least Squares (FGLS) estimation, the authors examined how broadband access disparities influence the effectiveness of citizen engagement in fostering accountability. The study contributes to the literature by offering a conceptual and empirical framework that treats the digital divide as a background condition and a dynamic factor shaping participatory outcomes. Operationalizing digital accountability through distinct dimensions of transparency also advances accountability measurement in the digital era. The findings underscore that expanding e-government services alone is insufficient; digital inclusion and civic engagement must be pursued together. The study highlights the importance of integrated strategies that combine infrastructure investment with efforts to ensure accessible, inclusive, and effective public participation. While focused on the EU, the framework holds relevance for other governance contexts addressing digital inequality.Publisher versio
Anomaly detection via graph contrastive learning
No full textGraph contrastive learning (GCL) techniques have shown superior performance in many tasks, such as social networks and recommendation systems, which makes them good solution candidates for detecting anomalies more accurately. The existing noncontrastive learning-based approaches do not fully take into account dynamic agents that can camouflage themselves. These agents either establish frequent associations with regular objects or intentionally skip forming relationships with the remaining objects, which are called head and tail anomalies respectively. In terms of graph topology, such agent behavior makes the graph more imbalanced. To handle both types of anomalies, we come up with a novel ensemble graph contrastive learning-based GCAD (Graph Contrasted Anomaly Detection) which is an ensemble of two approaches: 1- We learn representations and embeddings by leveraging Siamese architecture, which learns to minimize/maximize similarity between graph pairs at different scales while capturing their hierarchical structure. 2- We integrate a self-supervised learning framework using graph augmentations (like node and edge dropout) and contrastive learning to learn robust graph embeddings. In this case, the main idea is to generate multiple views of a graph using augmentations and then maximize the agreement between these views using contrastive loss. We show our approach outperforms the competing approaches in detecting both tail and head anomalies across 6 different datasets from citation and finance domains. The ablation studies also show the importance of GCAD components as well as its robustness
Relations between Turkey's national intelligence and foreign secret services until the early cold war
No full textThis article examines the Turkish intelligence service from its establishment to the early Cold War period, especially in activities surveiling the Soviet Union. Turkey pursued a cautious path in its intelligence gathering regarding the Soviet Union in the period between both World Wars, by cooperating with many countries, especially Nazi Germany after the Second World War broke out in 1939. This legacy formed the foundations of the relationships, which Turkish intelligence officials established with secret services in the West during the early phases of the Cold War
Investigation of bacterial cells and clays as rheology modifiers in 3D concrete printing
No full textThe developments in digital concrete production placed the concrete industry in a paradox for achieving optimization among improved performance, cost, and sustainability. The three-dimensional (3D) printing material design contradicts the sustainability goal of reducing raw material consumption. This study investigates the role of bacterial cells as rheology-modifying agents (RMA) to enhance the performance of cement-based mortars used in 3D concrete printing. Two bacterial species, Bacillus megaterium and Sporosarcina pasteurii, were incorporated with clays (nanomontmorillonite and sepiolite) in fly ash-amended mix designs. Rheological analysis demonstrated that the inclusion of bacterial cells reduced dynamic yield stress by up to 30%. Incorporation of cells with clays improved the development of static yield stress and increased thixotropy by a factor of ten compared to that of the control samples. The synergistic effect of the bacterial cells and clays further enhanced buildability, with nanomontmorillonite improving shape retention by 6%-9% relative to control mixes, whereas sepiolite contributed to better interlayer bonding. Modular prototypes with optimized mixes achieved layer heights that retained 92%-93% of the target dimensions, indicating improved geometric stability and surface finish. The results of this study will provide a better understanding of the influence of B. megaterium and S. pasteurii cells on the rheology of the fly-ash-amended clay containing printable cement-based mortars. This research provides insight into the potential of biobased RMAs to enhance extrusion, shape retention, and sustainability in 3D concrete printing, paving the way for scalable applications within the construction industry.TÜBİTA
Unlocking the value in product return data: Inventory management with sales dependent stochastic product return flows from multiple periods
No full textIn the fast fashion retail sector, handling product returns has become a significant challenge due to rapidly changing consumer preferences and high product return rates. These retailers are now inclined to consider product return flows in managing product inventories using detailed product return data. This study investigates an optimal inventory control policy for a retailer facing stochastic product returns from multiple previous sale periods to maximize expected profit during a single selling season. The problem is formulated using dynamic programming, and due to its computational complexity, we propose an Approximate Dynamic Programming value iteration algorithm using basis functions. Our proposed algorithm reduces the solution time drastically without a significant sacrifice from optimality. We quantify the value of leveraging detailed return information and demonstrate that our proposed model increases the retailer's profit by 9% in the base case and up to 31% considering other cases compared to a model ignoring such information, especially under decreasing product prices over time or per period order capacity constraints. Finally, using an extensive computational study, we propose managerial insights on how to best leverage the value in the product return data using advanced analytics for fast-fashion retailers.TÜBİTA
Measurement of the double-differential inclusive jet cross section in proton-proton collisions at √s=5.02 TeV
No full textThe inclusive jet cross section is measured as a function of jet transverse momentum pT and rapidity y. The measurement is performed using proton-proton collision data at s = 5.02 TeV, recorded by the CMS experiment at the LHC, corresponding to an integrated luminosity of 27.4 pb−1. The jets are reconstructed with the anti-kT algorithm using a distance parameter of R = 0.4, within the rapidity interval |y| < 2, and across the kinematic range 0.06 < pT< 1 TeV. The jet cross section is unfolded from detector to particle level using the determined jet response and resolution. The results are compared to predictions of perturbative quantum chromodynamics, calculated at both next-to-leading order and next-to-next-to-leading order. The predictions are corrected for nonperturbative effects, and presented for a variety of parton distribution functions and choices of the renormalization/factorization scales and the strong coupling αS.the Marie-Curie program and the European Research Council and Horizon 2020 Grant, contract Nos. 675440, 724704, 752730, 758316, 765710, 824093, and COST Action CA16108 (European Union); the Leventis Foundation; the Alfred P. Sloan Foundation; the Alexander von Humboldt Foundation; the Science Committee, project no. 22rl-037 (Armenia); the Belgian Federal Science Policy Office; the Fonds pour la Formation a la Recherche dans l'Industrie et dans l'Agriculture (FRIA-Belgium); the Agentschap voor Innovatie door Wetenschap en Technologie (IWT-Belgium); the F.R.S.-FNRS and FWO (Belgium) under the "Excellence of Science - EOS" - be.h project n. 30820817; the Beijing Municipal Science & Technology Commission, No. Z191100007219010 and Fundamental Research Funds for the Central Universities (China); the Ministry of Education, Youth and Sports (MEYS) of the Czech Republic; the Shota Rustaveli National Science Foundation, grant FR-22-985 (Georgia); the Deutsche Forschungsgemeinschaft (DFG), under Germany's Excellence Strategy - EXC 2121 "Quantum Universe" - 390833306, and under project number 400140256 - GRK2497; the Hellenic Foundation for Research and Innovation (HFRI), Project Number 2288 (Greece); the Hungarian Academy of Sciences, the New National Excellence Program - UNKP, the NKFIH research grants K 124845, K 124850, K 128713, K 128786, K 129058, K 131991, K 133046, K 138136, K 143460, K 143477, 2020-2.2.1-ED-2021-00181, and TKP2021-NKTA-64 (Hungary); the Council of Science and Industrial Research, India; ICSC - National Research Center for High Performance Computing, Big Data and Quantum Computing, funded by the EU NexGeneration program (Italy); the Latvian Council of Science; the Ministry of Education and Science, project no. 2022/WK/14, and the National Science Center, contracts Opus 2021/41/B/ST2/01369 and 2021/43/B/ST2/01552 (Poland); the Fundacao para a Ciencia e a Tecnologia, grant CEECIND/01334/2018 (Portugal); the National Priorities Research Program by Qatar National Research Fund; MCIN/AEI/10.13039/501100011033, ERDF "a way of making Europe", and the Programa Estatal de Fomento de la Investigacion Cientifica y Tecnica de Excelencia Maria de Maeztu, grant MDM-2017-0765 and Programa Severo Ochoa del Principado de Asturias (Spain); the Chulalongkorn Academic into Its 2nd Century Project Advancement Project, and the National Science, Research and Innovation Fund via the Program Management Unit for Human Resources & Institutional Development, Research and Innovation, grant B37G660013 (Thailand); the Kavli Foundation; the Nvidia Corporation; the SuperMicro Corporation; the Welch Foundation, contract C-1845; and the Weston Havens Foundation (U.S.A.).Publisher versio
Exploring the experiences of parents and young children during the COVID-19 outbreak
No full textYoung children and their parents faced difficulties in their daily functioning due to the ongoing COVID-19 pandemic. In this qualitative study, we explored the experiences and perceptions of parents and their young children regarding the pandemic process. A sample of 22 parents (mothers and fathers) and nine children aged between four and six years from the same families were interviewed. Thematic analysis was used to generate themes. These themes reflected changes in the daily life of the parents and children, increased household chores and childcare, conflict reconciliation within the family, and coping strategies employed by parents and children. These themes are discussed within the frameworks of Family Systems Theory and the Family Stress Model. Drawing on the theoretical underpinnings of Family Systems Theory and the Family Stress Model, we critically examined these themes to illuminate the intricate dynamics at play within family units amidst the pandemic upheaval. Through this exploration, we aimed to provide a comprehensive understanding of the pandemic's impact on the familial fabric and shed light on potential avenues for support
Biomedical applications of metal-organic frameworks revisited
No full textMetal-organic frameworks (MOFs) have been shown to be great alternatives to traditional porous materials in various chemical applications, and they have been very widely studied for biomedical applications in the past decade specifically for drug storage. After our review published in 2011 [Keskin and K & imath;z & imath;lel, Ind. Eng. Chem. Res. 2011, 50 (4), 1799-1812, 10.1021/ie101312k], we have witnessed a very fast growth not only in the number and variety of MOFs but also in their usage across a broad spectrum of biomedical fields. With the recent integration of molecular modeling and data science approaches to the experimental studies, biomedical applications of MOFs have been significantly accelerated positioning them as pivotal components in the regenerative medicine, medical imaging, and diagnostics. In this review, we visited the diverse biomedical applications of MOFs considering the recent experimental and computational efforts on drug storage and delivery, bioimaging, and biosensing. We focused on the underlying mechanisms governing the molecular interactions between MOFs and biological systems and discussed both the opportunities and challenges in the field to highlight the potential of MOFs in advanced therapeutics for cancer and neurological diseases.European Union (EU) European Commission Joint Research Centre ; TÜBİTAK ; Ministry of Science, Technological Development and Innovation of the Republic of SerbiaPublisher versio