Sever, Hayri
Loading...
Profile URL
Name Variants
Sever, H.
Job Title
Prof. Dr.
Email Address
sever@cankaya.edu.tr
Main Affiliation
Bilgisayar Mühendisliği
Status
Current Staff
Website
ORCID ID
Scopus Author ID
Turkish CoHE Profile ID
Google Scholar ID
WoS Researcher ID
Files
Sustainable Development Goals
3
GOOD HEALTH AND WELL-BEING
7
Research Products
16
PEACE, JUSTICE AND STRONG INSTITUTIONS
1
Research Products
This researcher does not have a Scopus ID.
This researcher does not have a WoS ID.
Scholarly Output
32
Articles
13
Views / Downloads
2463/1747
Supervised MSc Theses
3
Supervised PhD Theses
0
WoS Citation Count
65
Scopus Citation Count
115
WoS h-index
5
Scopus h-index
7
Patents
0
Projects
0
WoS Citations per Publication
2.03
Scopus Citations per Publication
3.59
Open Access Source
9
Supervised Theses
3
Google Analytics Visitor Traffic
| Journal | Count |
|---|
Current Page: 1 / NaN
Scopus Quartile Distribution
Competency Cloud
Scholarly Output Search Results
Now showing 1 - 10 of 32
Conference Object Sınıflandırmada Küçük ve Dengesiz Veri Kümesi Problemi(2019) Par, Öznur Esra; Akçapınar Sezer, Ebru; Sever, HayriVerilerinin sınıflandırılması, veri kümesinin küçük ve dengesiz olması durumunda zorlaşmakta ve sınıflama performansını direkt etkilemektedir. Veri setinin küçük olması ve/veya sınıflar arasında dengesizlik olması veri madenciliğinde büyük bir sorun haline gelmiştir. Sınıflama algoritmaları, veri setlerinin yeterli büyüklüğe sahip, dengeli olduğu varsayımı üzerine geliştirilmiştir. Bu algoritmaların çoğu, azınlık sınıfındaki örnekleri göz ardı ederken veya yanlış sınıflandırırken, çoğunluk sınıfa odaklanır. Medikal veri madenciliğinde bazı kısıtlardan dolayı küçük ve dengesiz veri seti problemi ile sıklıkla karşılaşılmaktadır. Çalışma kapsamında erişime açık hepatit veri seti, küçük veri setlerine bölünmüş, oluşturulan her bir veri seti uzaklık tabanlı yöntemlerle çoğaltılmıştır. Çoğaltılan veri setleri dört farklı makine öğrenmesi algoritması (Yapay Sinir Ağları, Destek Vektör Makineleri, Naive Bayes ve Karar Ağacı) kullanılarak sınıflandırılmış, elde edilen sınıflama sonuçları karşılaştırılmıştır.Conference Object Production and Retrieval of Rough Classes in Multi Relations(Ieee Computer Soc, 2007) Tolun, M.R.; Sever, H.; Gorur, A.K.Organizational memory in today's business world forms basis for organizational learning, which is the ability of an organization to gain insight and understanding from experience through experimentation, observation, analysis, and a willingness to examine both successes and failures. This basically requires consideration of different aspects of knowledge that may reside on top of a conventional information management system. Of them, representation, retrieval and production issues of meta patterns constitute to the main theme of this article. Particularly we are interested in a formal approach to handle rough concepts. We utilize rough classifiers to propose a preliminary framework based on minimal term sets with p-norms to extract meta patterns. We describe a relational rule induction approach, which is called rila. Experimental results are provided on the mutagenesis, and the KDD Cup 2001 genes data sets. © 2007 IEEE.Conference Object A Data Fusion Approach in Protein Homology Detection(2008) Sever, H.; Polatkan, A.C.; Oǧul, H.The discriminative framework for protein remote homology detection based on support vector machines (SVMs) is reconstructed by the fusion of sequence based features. In this respect, n-peptide compositions are partitioned and fed into separate SVMs. The SVM outputs are evaluated with different techniques and tested to discern their ability for SCOP protein super family classification on a common benchmarking set. It reveals that the fusion approach leads to an improvement in prediction accuracy with a remarkable gain on computer memory usage. © 2008 IEEE.Conference Object Citation - Scopus: 2Web Service-Based Turkish Automatic Speech Recognition Platform(Ieee, 2020) Polat, Huseyin; Sever, Hayri; Oyucu, SaadinIn response to the similar challenges in building large-scale distributed applications and platforms on the Web, microservice architecture has emerged and gained a lot of popularity in recent years. Therefore, both for the use of microservices and for the provided of the necessary interface for Automatic Speech Recognition (ASR), a web-based platform has been developed. Within firstly the scope of the study, a Turkish ASR system was developed. A web service structure was created to facilitate access to the ASR system. The access of methods and data in the web service structure was provided through Representational State Transfer (REST) web services and service layer. An interface was developed to enable interaction with the web service. The platform was developed using a combination of different technologies such as ASR, web services, microservices, and interface technologies. The developed platform can be used via a standard web browser or an Application Programming Interface (API). In this study, Docker packages were used to improve system performance instead of using different virtual machines on a single server. In the experiments performed, it was shown that the Turkish ASR system had a word error rate of 24.70%. In web service performance tests, it was shown that the platform responded in an average of 9.6 seconds for a 59-second speech recording. The developed user interface was tested in both mobile and desktop web browsers and was shown to function properly. Applications and other services were given access to the platform without the need to use an interface via API support provided by the platform. As a result, a web service-based Turkish ASR platform working seamlessly on the ever-increasing number of mobile devices, the Internet of Things ecosystem, or other access devices was developed.Conference Object Citation - WoS: 2Citation - Scopus: 15Securing Blockchain Shards by Using Learning Based Reputation and Verifiable Random Functions(Ieee, 2019) Ozsoy, Adnan; Sever, Hayri; Bugday, AhmetIn order to meet the increasing demand of the blockchain, it needs to find a solution to the scalability problem. It has been focused on sharding recently to address the scalability problem. In the sharding method, the blockchain is divided into pieces. Instead of a more extensive network, networks with fewer nodes are created. As a result, it becomes more important that each node in the network is reliable. In this study, studies using sharding method have been investigated, and methods for the assigning nodes to shards are proposed. The use of learning-based adaptive methods for this process will contribute to the safe and reliable use of shards. The probability of the shards to deteriorate and influence the whole blockchain will be reduced.Article Citation - WoS: 3Citation - Scopus: 3Binary Background Model With Geometric Mean for Author-Independent Authorship Verification(Sage Publications Ltd, 2023) Sezer, Ebru A.; Sever, Hayri; Canbay, PelinAuthorship verification (AV) is one of the main problems of authorship analysis and digital text forensics. The classical AV problem is to decide whether or not a particular author wrote the document in question. However, if there is one and relatively short document as the author's known document, the verification problem becomes more difficult than the classical AV and needs a generalised solution. Regarding to decide AV of the given two unlabeled documents (2D-AV), we proposed a system that provides an author-independent solution with the help of a Binary Background Model (BBM). The BBM is a supervised model that provides an informative background to distinguish document pairs written by the same or different authors. To evaluate the document pairs in one representation, we also proposed a new, simple and efficient document combination method based on the geometric mean of the stylometric features. We tested the performance of the proposed system for both author-dependent and author-independent AV cases. In addition, we introduced a new, well-defined, manually labelled Turkish blog corpus to be used in subsequent studies about authorship analysis. Using a publicly available English blog corpus for generating the BBM, the proposed system demonstrated an accuracy of over 90% from both trained and unseen authors' test sets. Furthermore, the proposed combination method and the system using the BBM with the English blog corpus were also evaluated with other genres, which were used in the international PAN AV competitions, and achieved promising results.Conference Object Sentiment Analysis for Arabic Using Deep Learning(Springer Science and Business Media Deutschland GmbH, 2026) al-Hamadani, S.A.S.; Sever, H.With the explosive growth of digital communication, understanding sentiment in online content has become increasingly critical for a wide range of applications, from customer feedback analysis to social media monitoring. However, sentiment analysis for Arabic presents unique challenges due to the language's rich morphology, diverse dialects, and complex syntactic structures. These challenges are further amplified in multimodal settings, where the fusion of textual, visual, and auditory cues is required to capture the full spectrum of human emotion. To address these issues, this paper introduces a new framework for Arabic Multimodal Sentiment Analysis (AMSA), combining multi-level deep learning approaches across text, audio, and visual modalities. Our approach utilizes state-of-the-art transformer-based architecturees, including Multimodal Transformer (MulT) and Early Fusion models, to tackle both linguistic complexity and multimodal alignment. Specifically, we leverage DeBERTa for extracting rich textual features, ViT (Vision Transformer) for visual cues, and Whisper for capturing nuanced audio signals, creating robust and contextualized representations. Experimental results on a curated Arabic multimodal dataset demonstrate the effectiveness of this approach, with our proposed MulT model achieving an F1 score of 72.73%, reflecting a substantial improvement of 13.98% in F1 score and 14.6% in accuracy over existing baselines. These findings highlight the power of cross-modal attention mechanisms and early fusion strategies in accurately capturing subtle sentiments across multiple modalities. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2026.Conference Object Authorship Modelling Approach for Authorship Verification on the Turkish Texts(Ieee, 2018) Akcapinar Sezer, Ebru; Sever, Hayri; Canbay, PelinAuthorship attribution which aims to extract information about an author by analyzing the text of the author is a challenging field that has been studied for years. This study becomes even more difficult when there is limited data on this field. The need for this study carried out under the name of Authorship Verification is increasing day by day with the increase of anonymous authors in the electronic environments. In this study, a model-based solution approach is presented for the authorship verification problem. With the presented approach, it was determined what should be the success interval to be considered in the authorship verification problem.Article Citation - WoS: 7Citation - Scopus: 13A Concept-Based Sentiment Analysis Approach for Arabic(Zarka Private Univ, 2020) Sever, Hayri; Nasser, AhmedConcept-Based Sentiment Analysis (CBSA) methods are considered to be more advanced and more accurate when it compared to ordinary Sentiment Analysis methods, because it has the ability of detecting the emotions that conveyed by multi-word expressions concepts in language. This paper presented a CBSA system for Arabic language which utilizes both of machine learning approaches and concept-based sentiment lexicon. For extracting concepts from Arabic, a rule-based concept extraction algorithm called semantic parser is proposed. Different types of feature extraction and representation techniques are experimented among the building prosses of the sentiment analysis model for the presented Arabic CBSA system. A comprehensive and comparative experiments using different types of classification methods and classifier fusion models, together with different combinations of our proposed feature sets, are used to evaluate and test the presented CBSA system. The experiment results showed that the best performance for the sentiment analysis model is achieved by combined Support Vector Machine-Logistic Regression (SVM-LR) model where it obtained a F-score value of 93.23% using the Concept-Based-Features + Lexicon-Based-Features + Word2vec-Features (CBF + LEX+ W2V) features combinations.Article Citation - WoS: 3Citation - Scopus: 4Block Size Analysis for Discrete Wavelet Watermarking and Embedding a Vector Image as a Watermark(Zarka Private Univ, 2019) Sever, Hayri; Sever, Hayri; Senol, Ahmet; Elbasi, Ersin; Bilgisayar MühendisliğiAs telecommunication and computer technologies proliferate, most data are stored and transferred in digital format. Content owners, therefore, are searching for new technologies to protect copyrighted products in digital form. Image watermarking emerged as a technique for protecting image copyrights. Early studies on image watermarking used the pixel domain whereas modern watermarking methods convert a pixel based image to another domain and embed a watermark in the transform domain. This study aims to use, Block Discrete Wavelet Transform (BDWT) as the transform domain for embedding and extracting watermarks. This study consists of 2 parts. The first part investigates the effect of dividing an image into non overlapping blocks and transforming each image block to a DWT domain, independently. Then, effect of block size on watermark success and, how it is related to block size, are analyzed. The second part investigates embedding a vector image logo as a watermark. Vector images consist of geometric objects such as lines, circles and splines. Unlike pixel-based images, vector images do not lose quality due to scaling. Vector watermarks deteriorate very easily if the watermarked image is processed, such as compression or filtering. Special care must be taken when the embedded watermark is a vector image, such as adjusting the watermark strength or distributing the watermark data into the image. The relative importance of watermark data must be taken into account. To the best of our knowledge this study is the first to use a vector image as a watermark embedded in a host image.
