OpenMETU (Middle East Technical University)
Not a member yet
110280 research outputs found
Sort by
Feature Distribution-Based Touch Biometrics Using CNN and Siamese Networks
No full textEnsuring secure and user-friendly authentication is important as mobile devices increasingly handle sensitive data. Traditional methods like PINs, fingerprints, and facial recognition have privacy limitations, whereas behavioral biometrics offer implicit, continuous authentication. This study presents a touch-based authentication framework leveraging feature distribution modeling with a Convolutional Neural Network (CNN)-based Siamese network. Using Kullback-Leibler (KL) divergence, we compare touch dynamics distributions across sessions to differentiate users. To address behavioral variability, we employ adaptive bandwidth tuning in kernel density estimation (KDE) for improved probability modeling. The CNN extracts embeddings from these feature distributions, while the Siamese network assesses session similarities. Unlike traditional handcrafted approaches using summary statistics, our method preserves the full statistical structure of touch interactions, improving authentication accuracy. Experimental results demonstrate competitive Equal Error Rates (EER), underscoring the potential of distribution-driven touch biometrics for mobile authentication
New data on text reading in English as a second language
Get PDFThis paper reports an expansion of the English as a second language (L2) component of the Multilingual Eye Movement Corpus (MECO L2), an international database of eye movements during text reading. While the previous Wave 1 of the MECO project (Kuperman et al., 2023) contained English as a L2 reading data from readers with 12 different first language (L1) backgrounds, the newly collected dataset adds eye-tracking data on English text reading from 13 distinct L1 backgrounds (N = 660) as well as participants' scores on component skills of English proficiency and information about their demographics and language background and use. The paper reports reliability estimates, descriptive statistics, and correlational analyses as means to validate the expansion dataset. Consistent with prior literature and the MECO Wave 1, trends in the MECO Wave 2 data include a weak correlation between reading comprehension and oculomotor measures of reading fluency and a greater L1-L2 contrast in reading fluency than reading comprehension. Jointly with Wave 1, the MECO project includes English reading data from more than 1,200 readers representing a diversity of native writing systems (logographic, abjad, abugida, and alphabetic) and 19 distinct L1 backgrounds. We provide multiple pointers to new venues of how L2 reading researchers can mine this rich publicly available dataset
Dataset for Modeling Anisotropy in 3D Printed Concrete
Get PDFThis dataset supports the study titled “A Semi-Empirical Framework for Modeling Anisotropy, Spatial Variation and Failure Mechanisms in 3D Printed Concrete”. It includes high-resolution computed tomography (CT) images, validated Abaqus input files for simulating cylindrical 3D-printed concrete (3DPC) cores, and a MATLAB-based CDP parameter generator for automated material calibration. The dataset enables reproducible finite element simulations that capture direction-dependent mechanical behavior and failure modes in 3DPC using a hybrid modeling approach. All resources have been structured to facilitate ease of use, promote transparency, and support further research on the mechanical modeling of additive manufacturing-based cementitious materials.TUBİTAK under grant agreement 120N990Research Universities Support Fund (ADEP-303-2022-11174
DeepKin: Predicting Relatedness From Low-Coverage Genomes and Palaeogenomes With Convolutional Neural Networks
No full textDeepKin is a novel tool designed to predict relatedness from genomic data using convolutional neural networks (CNNs). Traditional methods for estimating relatedness often struggle when genomic data is limited, as with palaeogenomes and degraded forensic samples. DeepKin addresses this challenge by leveraging two CNN models, which are trained solely on simulated genomic data, to classify relatedness up to the third degree and to identify parent–offspring and sibling pairs. Our benchmarking shows DeepKin performs comparably or better than the widely used tool READv2. We validated DeepKin, which uses PLINK's.map and.ped files as input, on empirical palaeogenomes from three archaeological sites, demonstrating its robustness and adaptability across different genetic backgrounds, with accuracy > 90% above 10 K shared SNPs. By capturing information across genomic segments, DeepKin offers a new methodological path for relatedness estimation in settings with highly degraded samples, with applications in ancient DNA, as well as forensic and conservation genetics
Brezilya Amazonlarında Ekstraktivizm, Ormansızlaşma ve Yerli Direnişi: Bolsonaro Sonrası Dönemde Ekolojik Adalet Arayışı
No full textYükselen Kore popülaritesi: Ankara'da Kore yeme-içme mekanları
No full textIn parallel with the global expansion of Korean cultural products through the phenomenon of the “Korean Wave” (Hallyu), K-pop (Korean music), K-dramas (Korean dramas), and the broader Korean entertainment industry have attracted significant attention. This trend has encouraged people to engage more deeply with Korean culture, and the consumption of products associated with Korean cuisine, known as K-food, has become an integral part of this expansion. The first Korean food and drink place in Ankara opened in 2018, and by 2025, this number had risen to seven. To understand this trend, this study focuses on seven Korean food and drink places in Ankara, examining them through the perspectives of both entrepreneurs and customers. While most studies on the Korean Wave in Turkey have focused on K-pop or K-dramas, this thesis highlights the localization of K-food as a new dimension of the Korean Wave in Ankara. Addressing this gap, the research draws on the theories of transnationalism and mixed embeddedness to shed light on how global cultural flows intersect with ethnic entrepreneurship, cultural diplomacy, and transnational networks. The study investigates why the number of Korean food and drink places in Ankara has steadily increased and why they have attracted such strong interest. To answer these questions, 13 semi-structured interviews were conducted with business owners/managers, and customers in Ankara in August 2024, complemented by participant observation. The findings reveal that ethnic entrepreneurship in this context has not emerged from migration-driven dynamics but rather as a result of Korean Wave influence; shaped by cultural curiosity, cultural diplomacy activities, media influence, and local market opportunities.“Kore Dalgası” (Hallyu) fenomeniyle birlikte Kore kültür ürünlerinin küresel yayılımına paralel olarak, K-pop (Kore müziği), K-dramalar (Kore dizileri) ve daha geniş anlamda Kore eğlence endüstrisi büyük ilgi görmüştür. Bu eğilim, insanların Kore kültürüyle daha yakından ilgilenmelerini teşvik etmiş ve K-food olarak bilinen Kore mutfağıyla ilişkili ürünlerin tüketimi, bu yayılmanın ayrılmaz bir parçası haline gelmiştir. Ankara'daki ilk Kore yeme-içme mekanı 2018 yılında açılmış ve 2025 yılında bu sayı yediye yükselmiştir. Bu eğilimi anlamak için, bu çalışma Ankara'daki yedi Kore yeme-içme mekanına hem girişimciler hem de müşteriler üzerinden odaklanmaktadır. Kore Dalgası ile ilgili çoğu çalışma Türkiye'de K-pop veya K-dramalar üzerine odaklanırken, bu tez Kore Dalgası'nın yeni bir boyutu olarak K-food'un Ankara’da yerelleşmesini vurgulamaktadır. Bu boşluğu ele alan bu araştırma, transnasyonalizm ve karma gömülülük teorilerinden yararlanarak, küresel kültürel akışların etnik girişimcilik, kültürel diplomasi ve transnasyonel ağlarla kesişimine ışık tutmaktadır. Çalışma, Ankara’daki Kore yeme-içme mekanlarının neden giderek arttığını ve neden yoğun bir ilgiyle karşılaştığını araştırmaktadır. Bu soruları yanıtlamak için, Ağustos 2024’te Ankara’da mekan sahipleri/işletmecileri ve müşteriler ile 13 yarı yapılandırılmış mülakat gerçekleştirilmiş ve bu mekanlara katılımcı gözlemde bulunulmuştur. Bulgular, etnik girişimciliğin göç kaynaklı dinamiklerden değil; Kore Dalgası’nın etkisiyle, kültürel merak, kültürel diplomasi faaliyetleri, medya etkisi ve yerel pazar fırsatlarının birleşiminden doğduğunu ortaya koymaktadır.M.S. - Master of Scienc
Exploring the Interrelationship of Emotion Regulation, Teacher Resilience, and Work Engagement among Pre- Service EFL Teachers in a Turkish University Context
No full textSections of rational elliptic Lefschetz fibrations
No full textWe give a list of monodromy factorizations in the pure mapping class group
of a torus with d+1 marked points that represent lines on a del
Pezzo surface Y of degree . These factorizations are lifts of a certain
fixed monodromy factorization in that represents Y. In the case d=1,
discussed in more detail, we give an explicit correspondence between such
factorizations and the 240 roots of , the orthogonal complement in
of the canonical class
A Seasonal Forecasting System for the Sacramento River Basin in California: Coupled Atmospheric-Hydrologic Numerical Model WEHY-HCM Integrated with a Lead Time-Dependent Seasonal Statistical Filternet
No full textTo forecast monthly river flows at Sacramento River Basin in California during February-July period of a year, a numerical atmospheric/hydrologic modeling system was coupled with a statistical updating system to create a seasonal flow forecasting system. This system first dynamically downscales global forecasts of Climate Forecast System Version 2 (CFSV2) to Sacramento River Basin at 9 km resolution by Weather Research Forecast (WRF) atmospheric model. Then, the refined climate forecasts from WRF are input into WEHY-HCM which is made up of the WEHY hydrology model, coupled to the WRF model, to develop monthly flow forecasts at various sub-basin outlets of Sacramento River during February-July period. These numerical model-based forecasts are then updated by a new seasonal exponential smoothing filter that is based on forecast lead times in order to issue the final monthly forecasts. This forecasting system is demonstrated in one of the subbasins of the Sacramento River
