检索结果-内蒙古大学图书馆

Who Follows Turkish Presidential Candidates in 2023 Elections?

学校读者我要写书评

暂无评论

Who Follows Turkish Presidential Candidates in 2023 Election...

IEEE Signal Processing and Communications Applications (SIU)

作者： Onur Varol Center of Excellence in Data Analytics Sabanci University

Social media serves as an important platform for political discussions, especially during election periods. While politicians use the platform to reach out to voters and share their campaigns, it also creates a debate environment for different voter groups. It is possible to see supportive and critical speeches simultaneously. Particularly as the election date approaches, undecided voters follow politicians on online platforms to learn about their campaigns. Changes in politicians' followers can be observed in detail by analyzing women and young voters who may be decisive in the results. In this study, we analyzed candidates for the 2023 Presidential elections of Türkiye and their followers on the Twitter platform. We examined how the demographic of these followers change over time. Prior to analyzing demographics such as gender and age group, we identified bot accounts and isolated them from the dataset. Significant events are reflected in the number of followers and the relationship between the observed anomalies and changes in the demographics have also been studied in this work.

关键词：

On the Convergence of Prey Predator Algorithm

学校读者我要写书评

暂无评论

International Journal of Applied and Computational Mathematics 2024年第2期10卷 1-23页

作者： Tilahun, Surafel Luleseged Department of Mathematics HPC and Big Data Analytics Center of Excellence Addis Ababa Science and Technology University Addis Ababa 16417 Ethiopia

Unlike their success in solving complex and difficult problems, there is a lack of proper mathematical analysis and discussion of metaheuristic algorithms. Even though some researches are done on these aspects, the gap is still wide and different researchers outlined that more needs to be done to understand and analyse the search behaviour mathematically in order to move towards explainable AI. Convergence analysis is one of the issues researchers are trying to address. In this paper, a discussion on the convergence of prey predator algorithm will be presented. It is a swarm based algorithm which shows promising results. Furthermore, some initial researches show that it is a generalized algorithm where some well known algorithms fall as its special case. Hence, its analysis can be customized and used for other algorithms as well. Based on mathematical arguments, this paper discusses the convergence analysis of the algorithm to a near optimal solution with appropriate implementation, i.e. appropriate parameter setting. Prey predator algorithm is shown to be an ϵ-convergent algorithm that means there is a non infinity iteration number to get at least epsilon less quality from an optimal solution. © The Author(s), under exclusive licence to Springer Nature India Private Limited 2024.

关键词： Convergence Metaheuristic Prey predator algorithm Swarm intelligence

Politics and Propaganda on Social Media: How Twitter and Meta Moderate State-Linked Information Operations

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Mugurtay, Nihat Duygu, Umut Varol, Onur Faculty of Engineering and Natural Sciences Sabanci University Turkey Sabanci University Faculty of Arts and Social Sciences Sabanci University Turkey Center of Excellence in Data Analytics Sabanci University Turkey

Why do Social Media Corporations (SMCs) engage in state-linked information operations? Social media can significantly influence the global political landscape, allowing governments and other political entities to engage in concerted information operations, shaping or manipulating domestic and foreign political agendas. In response to state-linked political manipulation tactics on social media, Twitter and Meta carried out take-down operations against propaganda networks, accusing them of interfering foreign elections, organizing disinformation campaigns, manipulating political debates and many other issues. This research investigates the two SMCs' policy orientation to explain which factors can affect these two companies' reaction against state-linked information operations. We find that good governance indicators such as democracy are significant elements of SMCs' country-focus. This article also examines whether Meta and Twitter's attention to political regime characteristics is influenced by international political alignments. This research illuminates recent trends in SMCs' take-down operations and illuminating interplay between geopolitics and domestic regime characteristics. Copyright © 2024, The Authors. All rights reserved.

关键词： Social networking (online)

Unsupervised detection of coordinated fake-follower campaigns on social media

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Zouzou, Yasser Varol, Onur Faculty of Engineering and Natural Sciences Sabanci University Turkey Center of Excellence in Data Analytics Sabanci University Turkey

Automated social media accounts, known as bots, are increasingly recognized as key tools for manipulative online activities. These activities can stem from coordination among several accounts and these automated campaigns can manipulate social network structure by following other accounts, amplifying their content, and posting messages to spam online discourse. In this study, we present a novel unsupervised detection method designed to target a specific category of malicious accounts designed to manipulate user metrics such as online popularity. Our framework identifies anomalous following patterns among all the followers of a social media account. Through the analysis of a large number of accounts on the Twitter platform (rebranded as Twitter after the acquisition of Elon Musk), we demonstrate that irregular following patterns are prevalent and are indicative of automated fake accounts. Notably, we find that these detected groups of anomalous followers exhibit consistent behavior across multiple accounts. This observation, combined with the computational efficiency of our proposed approach, makes it a valuable tool for investigating large-scale coordinated manipulation campaigns on social media platforms. © 2023, CC BY-NC-ND.

关键词： Computational efficiency

TurkishBERTweet: Fast and Reliable Large Language Model for Social Media Analysis

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Najafi, Ali Varol, Onur Faculty of Engineering and Natural Sciences Sabanci University Turkey Center of Excellence in Data Analytics Sabanci University Turkey

Turkish is one of the most popular languages in the world. Wide us of this language on social media platforms such as Twitter, Instagram, or Tiktok and strategic position of the country in the world politics makes it appealing for the social network researchers and industry. To address this need, we introduce TurkishBERTweet, the first large scale pre-trained language model for Turkish social media built using almost 900 million tweets. The model shares the same architecture as base BERT model with smaller input length, making TurkishBERTweet lighter than BERTurk and can have significantly lower inference time. We trained our model using the same approach for RoBERTa model and evaluated on two text classification tasks: Sentiment Classification and Hate Speech Detection. We demonstrate that TurkishBERTweet outperforms the other available alternatives on generalizability and its lower inference time gives significant advantage to process large-scale datasets. We also compared our models with the commercial OpenAI solutions in terms of cost and performance to demonstrate TurkishBERTweet is scalable and cost-effective solution. As part of our research, we released TurkishBERTweet and fine-tuned LoRA adapters for the mentioned tasks under the MIT License to facilitate future research and applications on Turkish social media. Our TurkishBERTweet model is available at: https://***/ViralLab/TurkishBERTweet. Copyright © 2023, The Authors. All rights reserved.

关键词： Cost effectiveness

From Occasional to Steady: Habit Formation Insights From a Comprehensive Fitness Study

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Demirci, Ege Tüzün, Efe Ün, Ahmet Furkan Sönmez, Taner Giray Varol, Onur Faculty of Engineering and Natural Sciences Sabanci University Istanbul34956 Turkey Mars Athletic Club Istanbul34340 Turkey Center of Excellence in Data Analytics Sabanci University Istanbul34956 Turkey

Exercising regularly is widely recognized as a cornerstone of health, yet the challenge of sustaining consistent exercise habits persists. Understanding the factors that influence the formation of these habits is crucial for developing effective interventions. This study utilizes data from Mars Athletic Club, Türkiye’s largest sports chain, to investigate the dynamics of gym attendance and habit formation. The general problem addressed by this study is identifying the critical periods and factors that contribute to the successful establishment of consistent exercise routines among gym-goers. Here we show that there are specific periods during which gym attendance is most crucial for habit formation. By developing a survival metric based on gym attendance patterns, we pinpoint these critical periods and segment members into distinct clusters based on their visit patterns. Our analysis reveals significant differences in how various subgroups respond to interventions, such as group classes, personal trainer sessions, and visiting different clubs. Using causal inference analysis, we demonstrate that personalized guidance and social dynamics are key drivers of sustained long-term engagement. By systematically examining these variables and considering the specific characteristics of different clusters, our research demonstrates the importance of a tailored, multi-dimensional approach to promoting exercise habits, which integrates social dynamics, personalized guidance, and strategic interventions to sustain long-term engagement. Copyright © 2025, The Authors. All rights reserved.

关键词： Cluster analysis

Novel Feature Extractor Framework in Conjunction with Supervised Three Class-XGBoost Algorithm for Osteosarcoma Detection from Whole Slide Medical Histopathology Images

学校读者我要写书评

暂无评论

Computers, Materials & Continua 2025年第2期82卷 3337-3353页

作者： Tanzila Saba Muhammad Mujahid Shaha Al-Otaibi Noor Ayesha Amjad Rehman Khan Artificial Intelligence&Data Analytics Lab College of Computer&Information SciencesPrince Sultan UniversityRiyadh11586Saudi Arabia Department of Information Systems College of Computer and Information SciencesPrincess Nourah Bint Abdulrahman UniversityRiyadh11671Saudi Arabia Center of Excellence in Cyber Security(CYBEX) Prince Sultan UniversityRiyadh11586Saudi Arabia

Osteosarcomas are malignant neoplasms derived from undifferentiated osteogenic mesenchymal cells. It causes severe and permanent damage to human tissue and has a high mortality rate. The condition has the capacity to occur in any bone;however, it often impacts long bones like the arms and legs. Prompt identification and prompt intervention are essential for augmenting patient longevity. However, the intricate composition and erratic placement of osteosarcoma provide difficulties for clinicians in accurately determining the scope of the afflicted area. There is a pressing requirement for developing an algorithm that can automatically detect bone tumors with tremendous accuracy. Therefore, in this study, we proposed a novel feature extractor framework associated with a supervised three-class XGBoost algorithm for the detection of osteosarcoma in whole slide histopathology images. This method allows for quicker and more effective data analysis. The first step involves preprocessing the imbalanced histopathology dataset, followed by augmentation and balancing utilizing two techniques: SMOTE and ADASYN. Next, a unique feature extraction framework is used to extract features, which are then inputted into the supervised three-class XGBoost algorithm for classification into three categories: non-tumor, viable tumor, and non-viable tumor. The experimental findings indicate that the proposed model exhibits superior efficiency, accuracy, and a more lightweight design in comparison to other current models for osteosarcoma detection.

关键词： Medical image processing deep learning healthcare image classification histopathology

SIU2023-NST - Hate Speech Detection Contest

学校读者我要写书评

暂无评论

SIU2023-NST - Hate Speech Detection Contest

IEEE Signal Processing and Communications Applications (SIU)

作者： İnanç Arın Zeynep Işık Seçilay Kutal Somaiyeh Dehghan Arzucan Özgür Berrin Yanikoğlu Faculty of Engineering and Natural Sciences Sabanci University Istanbul Turkey Center of Excellence in Data Analytics (VERIM) Sabanci University Istanbul Turkey Department of Computer Engineering Bogazici University Turkey

Hate speech spread on social media can strongly affect people and societies, especially when it targets a specific group of people in terms of religion, culture, or a specific social situation, such as refugees. For this reason, the detection and elimination of hate speech in social networks have attracted the attention of natural language processing researchers in recent years. A competition is organized to benchmark progress in Turkish hate speech recognition, within SIU2023 Computational Social Sciences special session, with four different tasks. A total of 20 teams registered for the competition, while eight teams submitted results at the end. The details of the winning models and their results are explained in this paper.

关键词：

Additive Manufacturing of Ti6Al4V Using Directed Energy Deposition: An Elaborative Approach of Process Modelling via Computational Fluid Dynamics

学校读者我要写书评

暂无评论

Lasers in Manufacturing and Materials Processing 2025年第1期12卷 201-225页

作者： Ur Rehman, Asif Azher, Kashif Rehman, Aqsa Shah, Mussadiq Mahmood, Muhammad Arif Tüfekci, Celal Sami Salamci, Metin Uymaz Department of Mechanical Engineering Gazi University Ankara Turkey Bartlett School of Sustainable Construction University College London LondonWC1E 6BT United Kingdom Department of Mechanical Engineering King Fahd University of Petroleum & Minerals Dhahran31261 Saudi Arabia Data Science and Analytics Institute The University of Oklahoma NormanOK73019 United States Institute of Chemical Technologies and Analytics TU Wien Getreidemarkt 9 Vienna1060 Austria Intelligent Systems Center Missouri University of Science and Technology RollaMO65409 United States Advanced Manufacturing Technologies Center of Excellence-URTEMM Ankara Turkey Additive Manufacturing Technologies Application and Research Center-EKTAM Ankara Turkey

Additive manufacturing (AM) has revolutionized many industries. Directed energy deposition (DED) offers remarkable utility in size and is an effective method in terms of energy efficiency. The interaction of powder with a laser before coming in contact with the substrate helps to attain fair melting and deposition. However, there is a problem associated with non-uniformity and accumulation of metallic melt on the edges of the manufactured parts which needs to be understood. Herein, the DED process of Ti6Al4V alloy is presented. The phenomenon of recoil pressure and Marangoni actions along the single tracks of the melt pool is defined via the computational fluid dynamics (CFD) analysis. The discrete element modeling (DEM) improves the simulation model output as variable particle diameters are considered, thus mimicking the actual powder size distribution. The simulation results disclosed intricate facts about the flow dynamics of molten material under varying conditions. The stream traces and thermal gradients are evaluated to recognize the effect of laser movement on material cooling and surface tension. The material accumulation occurred at curves and borders due to cooling-induced surface tension alterations, affecting the attribute of the deposited layers. The bi-directional flow patterns are prompted by Marangoni convection and recoil pressure, causing distinct features, including lateral menisci and depression modes. The cooling rates impacted the liquid portion and dilution levels, critical for attaining optimal bonding between deposited layers and substrate. Experimental endorsement using Synchrotron X-ray imaging favored the simulation results, confirming the simulated Marangoni convection effects. This comparison highlights the efficiency of CFD models in predicting molten material behavior, offering a worthy tool for adjusting DED printing parameters. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature

关键词： Titanium alloys