Faculté des mathématiques et de l'informatique
Permanent URI for this communityhdl:123456789/17
Browse
483 results
Search Results
Item Profilage hybride et mise à jour adaptative des profils utilisateur pour la recommandation Personnalisée dans les plateformes e-Learning(university of bordj bou arreridj, 2025) Bouafia Amani; Charifi KarimaLes systèmes de recommandation ont énormément amélioré l’expérience utilisateur sur internet. En particulier, les systèmes de recommandation dans l’e-Learning ont joué un rôle clé en aidant les étudiants à découvrir de nouveaux cours pertinents, basés sur des facteurs spécifiques et leur comportement sur la plateforme. Notre objectif principal est de créer un système de recommandation hybride en combinant deux modèles : l’approche basée sur le contenu et l’approche de filtrage collaboratif. Le principal problème pour les étudiants lorsqu’ils étudient en ligne est qu’ils sont exposés à une grande quantité de données qui peut nuire à leur réussite académique notre algorithme améliore l’apprentissage automatique, et les résultats démontrent son efficacité en termes de qualité et de pertinence .L’apprentissage automatique aidera notre système à comprendre le comportement des étudiants grâce a de nombreuses méthodes ce qui permettra d’obtenir des informations sur le contenu le plus susceptible de leur être pertinent, cette recherche constitue une modeste contribution au domaine des systèmes de recommandation et met en lumière leur potentiel à améliorer l’expérience et la productivité des étudiants.Item Eye tracking in simple visual search tasks(university of bordj bou arreridj, 2025) BOUHADDA KENZA; BOUDIAF FELLAEye tracking has become an essential technique for understanding human visual attention and behavior across a wide range of fields. One of the core challenges in this domain is ac curately predicting gaze during visual search tasks. As traditional models using handcrafted features often lack generalizability,Recent advances in deep learning offers a powerful alterna tives by enabling data-driven learning of complex spatial patterns in visual attention. This research introduces a deep learning-based eye-tracking system aimed at predicting vi sual attention through saliency maps, using the uEyes dataset which features a variety of image categories including desktop, mobile, web, and posters, along with corresponding human eye tracking data . the system employs a U-Net convolutional neural network optimized for pixel level saliency prediction. The model is trained and evaluated using a robust set of performance metrics, including Mean Absolute Error (MAE), Mean Squared Error (MSE), Kullback-Leibler Divergence (KLD), Correlation Coefficient (CC), Histogram Similarity (SIM), and Accuracy. This system showcases the effectiveness of deep learning in modeling human visual behav ior in visual search tasks.Item Détection automatique des faux comptes sur les réseaux sociaux à l’aide de l’apprentissage profond(university of bordj bou arreridj, 2025) Boumaiza Amdjad; Mohammedi ChahinezIn the digital age, social media platforms have become essential tools for communication, information dissemination, and brand visibility. However, this widespread use has given rise to a growing concern: the proliferation of fake accounts, particularly on Instagram. These inauthentic profiles, often automated or maliciously crafted, pose serious threats to user security, distort engagement metrics, and serve as vehicles for disinformation and fraudulent activities. To address this challenge, this thesis presents a deep learning-based approach using Long Short-Term Memory (LSTM) neural networks, which are well-suited to modeling the sequential and behavioral data of social media users. A synthetic dataset representing Instagram accounts was used to train and evaluate the model. The results highlight the method’s ability to accurately classify accounts as genuine or fake, offering strong performance metrics and promising generalization capabilities. This research contributes to the broader field of cybersecurity and illustrates the potential of artificial intelligence in detecting online threats and enhancing digital platform integrityItem Machine Learning for Misbehavior Detection in Next-Generation Vehicular Networks(university of bordj bou arreridj, 2025) MADI Ahmed Salah Eddine; MEKHFI BayaConnected vehicles have great potential to enhance road safety, reduce traffic congestion, and play a vital role in green engineering by reducing pollution and fuel consumption. By enabling more efficient traffic flow, eco-routing, and optimized driving behaviors, connected vehicles contribute to a cleaner environment. However, when a vehicle is compromised, it can pose a serious threat to the entire network due to the potential harm it can cause. One of the major challenges in vehicular networks is the detection of misbehaving vehicles, which should then be blacklisted or their certificates revoked. In this work, we propose a novel scheme that leverages machine learning to accurately detect and classify vehicle behavior, enabling effective identification and management of misbehaving vehicles. To assess the effectiveness of our approach, a comprehensive comparative analysis was performed. The results demonstrate that our model outperforms existing methods in accurately classifying vehicle behaviors, highlighting its potential for real-world deployment in securing vehicular networks.Item Autism Spectrum Disorder Detection Using Deeplearning Techniques(university of bordj bou arreridj, 2025) Benyzid Slimane; Fareh SamirThis project aims to develop an intelligent system for detecting autism spectrum disorder (ASD) using deep learning techniques. Autism is a complex condition that affects communica tion and behavior, and early diagnosis is critical for effective, appropriate, and prompt interven tion. The system uses convolutional neural networks (CNNs), such as MobileNet and VGG19, to classify individuals as having or not having autism based on face images and eye-tracking data. Apublicly available Kaggle dataset containing images representing typical visual behavior of individuals with ASD was used. The data was preprocessed through resizing, normalization, and augmentation to improve model performance. The model was evaluated using precision, accuracy, recall, F1 score, and ROC-AUC. Theresults demonstrated high performance and outperformed traditional methods, demons trating the model’s effectiveness in detecting autism. This project highlights the role of artificial intelligence in advancing healthcare by enabling faster and more accurate diagnosis of complex conditionsItem Catégorisation multi-étiquette des textes arabes(university of bordj bou arreridj, 2025) Bendib Yassamine; Bouaziz MeriemCe mémoire porte sur la classification multi-étiquettes des textes arabes à l’aide de techniques d’apprentissage profond. La langue arabe présente plusieurs défis, notamment une complexité morphologique élevée, une richesse lexicale importante et une grande variabilité dialectale. Ce travail se concentre sur l’utilisation de réseaux de neurones profonds, en particulier le modèle convolutionnel (CNN), BiLSTM et le modèle pré-entraîné AraBERT, afin d’améliorer la précision du classement. Le mémoire aborde les étapes clés du traitement automatique de la langue arabe, les méthodes de représentation des textes, ainsi que les métriques d’évaluation adaptées à la classification multi-étiquettes. Une attention particulière est portée à l’apprentissage par transfert, qui permet de tirer parti de la puissance des modèles linguistiques pré-entraînés pour la langue arabeItem Design and Implementation of an Innovative Digital Platform for Vehicle Breakdown Assistance(university of bordj bou arreridj, 2025) Oussama RAHIM; Mohammed salah eddine SAIDANIThis project introduces a digital platform to transform vehicle breakdown assistance in Al geria, addressing inefficiencies in traditional roadside services. It comprises three mobile appli cations for stranded motorists, autonomous repairmen, and workshop repairmen, and two web interfaces for workshop and platform administrators, built using Flutter, Next.js, Node.js with Express.js, and Firebase Realtime Database. The platform ensures scalability and real-time synchronization, offering features like GPS-based technician matching, secure user registra tion, a cash-based payment system, and performance analytics. Targeting motorists, repairmen, and administrators, it streamlines service requests and operational management. Analysis of existing solutions like DZ Dépannage revealed gaps in real-time tracking and payment systems, which this platform addresses with a user-centric design. The platform was developed using the Scrum framework across six sprints, utilizing a NoSQL database for flexible data management and efficient geolocation-based service provision.Item Intégration de l’Apprentissage Profond pour l’Optimisation du Routage dans les Réseaux FANETs(university of bordj bou arreridj, 2025) - KHIER Dounia; BELMOUMENE HoudaThis graduation thesis falls within the domain of mobile wireless networks, spe cifically Flying Ad Hoc Networks (FANETs), and proposes an intelligent approach to improve the stability of drone communications. The main objective is to design, mo del, and evaluate a predictive routing protocol capable of distinguishing stable links from unstable ones, in order to ensure better Quality of Service (QoS).The work be gins with an in-depth study of drone architectures in FANETs, followed by a critical analysis of conventional routing protocols and QoS management methods. Based on this analysis, we proposed an enhancement of the P-OLSR protocol by integrating a model based on Machine Learning and Deep Learning techniques to predict the Ex pected Transmission Count (ETX) metric, a key indicator of link quality.The proposed model was trained and validated using a dataset simulating FANET link characteris tics. A series of simulations was conducted to evaluate the performance of our protocol in terms of connection stability, throughput, latency, and routing reliability.The final performance of the model on the test set shows an accuracy of 98.58%, with a pre cision of 98.37%, a recall of 98.38%, and a low loss of 0.0612, demonstrating strong generalization ability.Simulation results also confirm the superiority of the proposed protocol, with a stability rate of 98.8%%, a reduced loss of 3%, a useful transmission rate of 97%, a resilience of 6%, and 92% active links, thus outperforming OLSR and Predictive-Optimized Link State Routing (P-OLSR).Item An Optimized Protocol for Mobile Sink Position Dissemination in Wireless Sensor Networks(university of bordj bou arreridj, 2025) Yasmine HAMMACHE; Aya Hiba BERGHEULIn wireless sensor networks, the mobility of the sink improves data collection efficiency; however, it presents new issues as sensor nodes must be aware of the mobile sink’s current loca tion. The dissemination of this location via flooding results in increased energy consumption. This study presents a new protocol called An Optimized Protocol for Mobile Sink Position Dissemination in Wireless Sensor Networks (OPMPD), which is specifically designed for sce narios involving random movement of the sink and event-based data collection. The proposed solution seeks to address the aforementioned problem, wherein specific chosen nodes retain the sink’s latest position while decreasing energy consumption and controlling overhead. Using the NS-3 simulator, the performance of OPMPD is assessed, and the results, depicted through graphs, demonstrate its superiority over the comparison protocol known as VGB across various metrics achieving 63.25 % reduction in energy consumption, 47,52 % improvement in network lifetime, and 86.17 % reduction in control overhead.Item Classification des Signaux EEG basée sur l’Apprentissage automatique et l'apprentissage profond(university of bordj bou arreridj, 2025) Temhachet Rayane; Laichaoui YasminaApproximately 50 million people worldwide suffer from epilepsy, a chronic neurological disorder. The automatic detection of epileptic seizures from EEG (electroencephalogram) signals remains a major challenge for researchers. This study proposes a methodology for classifying EEG signals using machine learning techniques, including Support Vector Machines (SVM), k-Nearest Neighbors (KNN), the Naive Bayes (NB) algorithm, and a Convolutional Neural Network (CNN). To evaluate model performance, metrics such as Accuracy (AC), Sensitivity (SE), Specificity (SP), and Receiver Operating Characteristic (ROC) curves were used. The results demonstrate the effectiveness of the applied models, particularly the CNN, in the automated detection of epilepsy from EEG signals.