diff --git "a/assets/projects_clean.csv" "b/assets/projects_clean.csv"
new file mode 100644--- /dev/null
+++ "b/assets/projects_clean.csv"
@@ -0,0 +1,104 @@
+Project_ID,Title,Abstract,Department,Year,Keywords,Source_Page
+P001,Automatic Plant Watering and Tracking System,"Many people struggle with watering their plants regularly, which usually leads to plants becoming unhealthy or dying. To address this common issue, it is important to monitor basic factors such as soil moisture and temperature in order to determine the proper time and amount of water needed. Since manual watering is not always consistent or reliable, this automatic plant watering system is designed to assist gardeners by providing an accurate irrigation based on real time weather conditions and the specific water requirements of each plant. The system includes a monitoring mechanism that helps ensure proper and efficient water usage, reducing water waste and supporting healthy plant growth. Additionally, the application includes a built-in encyclopedia that offers detailed information about a variety of plant types, helping users to understand their plant’s water needs, caring details and other useful facts. This solution makes it easier to take care for plants, especially when they are located away from owners.",Computer Science,2024 – 2025,"Moisturem Air humidity, AWTS system, Waterin, Plants .",5
+P002,علامة CourseAssist: CLO Performance Assistant “E-CPA”,"The rapid evolution of educational technology necessitates innovative approaches to enhance the assessment of student performance and course effectiveness. Our project addresses a significant issue present when faculty members spend an excessive amount of time completing the necessary documentation to achieve course objectives. To tackle this problem, we developed ""Alama,"" an automated mobile application that streamlines the traditional Excel data entry process. This application calculates Course Learning Outcomes (CLOs) and Program Learning Outcomes (PLOs) automatically, generating performance reports in moments rather than relying on manual entry. Alama not only mitigates the inefficiencies of conventional assessment methods, which can lead to inaccuracies and impede effective decision-making, but also integrates both quantitative and qualitative data. This holistic approach enhances the evaluation of educational processes, ultimately improving the learning experience for students. In conclusion, the implementation of Alama presents a significant advancement in academic assessment, offering both advantages—such as time savings and enhanced accuracy—and potential drawbacks, including the initial learning curve for faculty. Overall, Alama holds the promise of transforming educational assessment practices, leading to better educational outcomes.",Computer Science,2024 – 2025,"Program Learning Outcomes, Course Learning Outcomes, Performance Assistant, Formative, Summative",6
+P003,واصف Mobile Application for Deaf and Mute individuals. (WASIF),"Our System, Wasif, offers a communication application for deaf and mute individuals, we aim to develop cross-platform application that bridges these communication gaps by translating spoken language into sign language symbols and written text, while converting sign language gestures into audible speech. Wasif application analyzes hand and facial movements in real-time, ensuring accurate and efficient communication. Built using robust frameworks which is TensorFlow, and OpenCV, seeks to empower deaf and mute individuals to fully engage in all aspects of life, it is have been validated through rigorous testing with target user groups, ensuring that it meets the real-world needs.",Computer Science,2024 – 2025,"Arabic Sign Language, Sign Language Gestures into Audial Speech, Real-time Facial and Hand Movement Tracking, Deaf and mute.",7
+P004,أنيس Pet Care Application System (Anees),"Pets have always played a vital role in human companionship, offering emotional support, reducing stress, and enhancing overall well-being. As pet ownership continues to grow in Saudi Arabia, so there is a high demand for accessible, high-quality pet care services. However, many pet owners struggle to find reliable solutions for essential needs such as premium pet food, veterinary care, and grooming. Despite the increasing demand, there remains a significant gap in mobile applications dedicated to pet care in the region. To better understand this gap, extensive background research was conducted, including a comparison of existing pet care applications and, a comprehensive online questionnaire was distributed to residents of Saudi Arabia, targeting pet owners to assess their challenges and interest around a centralized pet care platform. The results highlighted a strong demand for a dedicated app that simplifies the pets owners lives by integrating essential services into a single, user-friendly platform. In response, Anees was conceptualized as an all-in-one solution, offering a pet food subscription service, appointment booking for veterinary and grooming needs, and an AI based breed identification feature using image recognition. Additionally, Anees includes a built-in marketplace for pet-related products, ensuring pet owners have seamless access to their pets supplies. By prioritizing usability, security, and reliability, Anees aims to revolutionize the pet care experience in Saudi Arabia, encouraging more individuals to adopt pets while providing existing owners with convenient, high-quality care.",Computer Science,2024 – 2025,"Pet care, mobile application, Artificial Intelligence, breed identification.",8
+P005,Thabat – ثبات,"Parkinson’s disease is a progressive neurodegenerative disorder that severely impacts motor functions, cognition, speech, and hand dexterity. Traditional therapy sessions for managing Parkinson’s disease symptoms often require in-person visits, which may be challenging for elderly patients due to mobility limitations and lack of consistent support. To address these challenges, we developed Thabat, a mobile application designed to offer virtual functional and cognitive rehabilitation specifically tailored for Arabic-speaking Parkinson’s patients. Thabat provides a suite of exercises across four essential therapeutic domains: physical, cognitive, speech, and hand dexterity. It leverages modern technologies such as Google ML Kit’s Pose Detection and speech recognition to monitor user performance and deliver feedback. These features enable patients to perform guided therapy from the comfort of their homes, supported by instant motivational cues. Thabat is built using Flutter and supports only the Android platform, ensuring accessibility for a wide user base in the target demographic. What distinguishes Thabat from existing solutions is its all-in-one approach, Arabic language support, and cost-free availability. By integrating body tracking, motivational support, and feedback in an easy-to-use interface, Thabat aims to improve therapy adherence and the overall quality of life for Parkinson’s disease patients. Furthermore, the app's reward-based system is designed to stimulate dopamine release, encouraging continued engagement. Thabat stands as an innovative contribution in the realm of mobile Health applications, offering a culturally sensitive and technologically advanced alternative for Parkinson’s disease rehabilitation.",Computer Science,2024 – 2025,"Parkinson's Disease, Rehabilitation, Pose Detection, Voice Recognition.",9
+P006,YUSUR,"Foreign pilgrims face several challenges while performing Hajj rituals. Many lack knowledge about how to perform the rituals or recite the specific prayers for Hajj, which might lead them to unreliable or incorrect sources. Additionally, they may experience getting lost during the rituals, losing contact with their assigned campaign, or struggling to communicate with campaign admin. The “Yusur” app serves as a comprehensive solution to meet pilgrims’ needs. It is a comprehensive application designed to assist Muslims worldwide in performing Hajj in Mecca. The app provides reliable and accurate information and supports the English language, making it easier for pilgrims to understand and correctly follow the rituals, enhancing their spiritual experience while reducing anxiety and confusion. The app offers trustworthy educational resources about Hajj rituals, including prayer times, Quran recitation, and key fatwas related to Hajj. It also features a map for locating the Hajj ritual sites and an option to request medical assistance when needed. All these services are available for pilgrims, certain functions require the pilgrim to be registered with a specific campaign. For Hajj campaign, “Yusur” provides innovative solutions, such as the ability to easily add their pilgrims, input essential information about them, and remove a pilgrim if necessary. Additionally, the app facilitates communication between campaign admin and pilgrims, helping them coordinate schedules, set meeting points, and efficiently organize plans. By integrating technology to meet the needs of pilgrims and campaign admin, the “Yusur” app aims to address common issues faced by foreign pilgrims. It enhances their safety and Hajj experience by combining all essential features that cater to the needs of both pilgrims and campaign admin, ensuring comfort, well-being, and safety for everyone.",Computer Science,2024 – 2025,"Hajj, Islamic app, Tasks organizer",10
+P007,Anees,"Anees is a mental health chatbot that supports users in both Arabic and English by offering helpful conversations, self-care tips, and emotional support. It uses well-known psychological methods like Cognitive Behavioral Therapy (CBT), Acceptance and Commitment Therapy (ACT), and encourages physical activity to promote better mental health. Anees is built using Flask for the backend and Flutter for the app interface, making it work smoothly on both iOS and Android. It combines rule-based logic with powerful AI language models to understand what users are feeling and respond in an empathetic way. The chatbot also includes a mood tracking feature that helps users see their emotional patterns over time. Anees focuses on making mental health support more accessible, especially for Arabic speakers, while also protecting user privacy and promoting ethical AI use. Early testing shows that Anees offers meaningful and supportive chats, making it a useful tool for improving emotional well-being in different communities",Computer Science,2024 – 2025,"Mental Health, Chatbot, Arabic Language, Emotional Support, AI Therapy",11
+P008,SyncUp,"Remote teamwork has revolutionized project implementation, offering a flexible and collaborative solution to modern workplace demands. However, the abundance of scattered applications for meetings, task management, and communication often cause inefficiencies and distraction. To address these challenges, we propose SyncUp, a unified platform that enables seamless project management and real-time collaboration. SyncUp allows users to create and manage projects, edit shared files collaboratively, leave notes, and track progress using intuitive visual indicators that highlight remaining time and tasks. Additionally, users can seek expert advice when needed, fostering a community-driven approach to problem-solving. The platform's design incorporates agile methodologies, ensuring iterative improvements based on user feedback and evaluations. By integrating these features, SyncUp aims to streamline remote teamwork, enhance productivity, and promote effective collaboration.",Computer Science,2024 – 2025,"Remote collaboration, project management, real-time teamwork, agile development, collaborative platform.",12
+P009,Apply: Streamlining Application Processes and Enhancing Career Development,"Apply is an innovative and inclusive platform designed to streamline the job and internship application process for students and job seekers. It integrates powerful features such as document organization, autofill for repetitive form fields, personalized job recommendations, application status tracking, and an AI-powered assistant (Octu-pal) that optimizes resumes, proofreads essays, and prepares users for interviews. The platform is developed using Swift in Xcode, targeting iOS devices, and includes seamless app extensions like a Safari autofill tool and widgets to enhance accessibility and user productivity. Apply addresses the often fragmented and overwhelming nature of the application process by centralizing essential tasks into one cohesive ecosystem. It reduces cognitive load, saves time, and boosts user confidence, ultimately increasing their chances of success. The platform also emphasizes user-centered design; it offers a seamless and intuitive user experience, ensuring that every step of the application process feels clear, personalized, and efficient. Through this streamlined approach, Apply empowers users to take control of their professional journeys with clarity and efficiency. As a future startup, Apply envisions contributing to a more equitable and accessible job market, where opportunities are easier to reach, better tailored, and more inclusive than ever before.",Computer Science,2024 – 2025,"AI Assistant, Autofill, Safari Extension, Streamlined Application Process, User-Centered Design",13
+P010,Dexter,"With artificial intelligence becoming increasingly accessible, distinguishing between human-written and AI-generated text has become more challenging. A lot of people nowadays rely on AI to generate content for everyday tasks, including expressing personal thoughts and feelings. This increasing reliance may result in issues like spam, false information, and misleading content. This graduation project introduces “Dexter”, an AI detection tool developed to detect a text that is produced by an AI and promote transparency in digital communication. Dexter enables users to input any text and uses advanced algorithms to analyze and determine whether the content is human-written or AI-generated. The system presents results clearly, displaying percentages and visually highlighting the segments most likely written by an ai. It also provides visual analysis through a pie chart, offering users intuitive and immediate feedback. Ultimately, Dexter’s mission is to empower users to recognize and differentiate between authentic human expression and AI-generated content, fostering trust and authenticity in digital interactions.",Computer Science,2024 – 2025,"AI-generated, AI detection, Dexter.",14
+P011,University Community App,"University students often face challenges in communication, social engagement, and accessing essential academic or university related information. Traditional tools like email, posters, or scattered messaging platforms fail to support students' evolving needs. This project introduces UniHub, a modern mobile application designed specifically for university communities to improve communication, collaboration, and student participation in campus life. UniHub acts as a centralized platform where students can connect with peers, post updates, join conversations, and stay informed about student clubs and university events. Key features include real-time chat (both group and direct), a section to connect with potential teammates for projects or competitions, and a course rating system that allows students to give and view feedback on courses. Students can also earn scores by posting within the app, encouraging engagement and contribution. The app also includes a ""Moments"" section where users can share highlights of their university experience from achievements to personal reflections promoting a stronger sense of connection and community. UniHub was developed based on direct feedback collected through student surveys and interviews, ensuring it reflects real needs. Built using Flutter and Laravel, the app provides a smooth, responsive experience across Android and iOS platforms. It blends familiar social networking tools with features tailored to university environments, helping students stay engaged, informed, and connected. By bridging communication gaps and bringing all student related content into one user friendly app, UniHub enhances both academic and social aspects of university life, making it easier for students to thrive in their studies and connect with their campus community.",Computer Science,2024 – 2025,"Student engagement, University platform, Social collaboration, Academic networking",15
+P012,خلها علينا ‘Khalaha Alyna’ – Event Management App,"This project aims to develop Khalaha Alyna, a user-friendly mobile application for event planning and management, designed for users in Saudi Arabia. The app streamlines event organization with features like event creation, guest list management, invitation distribution, and vendor coordination. Using the Waterfall methodology, the project follows a structured process from requirement gathering to implementation. Today’s methods, such as manual planning or complex software, are time-consuming and prone to errors—especially when managing guest lists and schedules. Khalaha Alyna addresses these challenges by offering a simple, centralized solution for both personal and professional events. The application is designed for individuals, event planners, and small businesses who often struggle with limited access to efficient planning tools. By focusing on a clear and intuitive user interface, the app improves the user experience and minimizes the need for support. The goal is to simplify event management and replace disorganized, complex methods with a smooth and hassle-free planning process.",Computer Science,2024 – 2025,"Event Planning, Mobile Application, Khalaha Alyna, User-Friendly, Vendor Coordination.",16
+P013,BusMap,"The BusMap mobile application is an innovative solution aimed at transforming public transportation through the integration of modern technologies such as Artificial Intelligence (AI) and real-time data tracking. The primary objective of the app is to enhance the efficiency, safety, and convenience of daily commuting for both passengers and transportation operators. Using AI algorithms, BusMap intelligently optimizes bus routes by arranging student or passenger stops in the most logical and time-efficient order based on proximity and traffic patterns. This ensures that bus drivers follow the fastest route possible while minimizing delays and fuel consumption. In addition to route optimization, the app provides real-time updates on bus locations, estimated arrival times, and any sudden changes or disruptions in the route. This transparency allows passengers to better plan their journeys and avoid long waiting times. For transportation operators and school supervisors, BusMap offers tools for performance tracking, shift scheduling, and attendance management, improving oversight and accountability across the system. By combining user-friendly design with cutting-edge backend logic, BusMap not only enhances the overall commuting experience but also contributes to smarter, greener urban transportation. The app ultimately aims to reduce operational costs, improve safety, and increase reliability for all stakeholders involved.",Computer Science,2024 – 2025,"location-based Services, traffic management, public transporting.",17
+P014,Integrated Sustainable Marketplace and Support Platform for Local Farmers (Rooted),"Rooted is an app that provides a virtual farmers market that helps small farmers promote and showcase their locally sourced produce or verified organic products directly to environmentally conscious consumers, providing farmers with jobs and additional income opportunities. Rooted also offers a specialized agricultural AI assistant, that gives personalized guidance and advice based on the farmers’ struggles growing their crops. Small local farmers in Saudi Arabia face many challenges marketing their organic or local products directly to consumers. Challenges include overflowing inventory, middlemen taking a high percentage of their income, and the lack of experience in selling and promoting their products. All issues listed above are mentioned in the National Agriculture Strategy Executive Summary for the year 2030. This demonstrates how important these problems are and that we quickly need to solve them. Our aim is to solve these problems by connecting small local farmers with environmentally conscious consumers. The app will showcase farmers' products, provide guidance to the farmer, and offer online ordering. Additionally, the app will be available in Arabic and English, as well as giving the option of delivery or farm pickup. Rooted aims to promote sustainable agriculture practices, reduce food waste, and provide job opportunities for small farmers.",Computer Science,2024 – 2025,"AI Chatbot, Agriculture, Sustainability, Digital Marketplace, Locally-Sourced",18
+P015,Date Harvesting Robot,"Traditional date harvesting is labor-intensive and inefficient, leading to losses and quality issues. This project introduces an AI-powered robotic system that automates date harvesting using computer vision, LiDAR sensors, and a robotic arm with a suction mechanism. The robot detects, classifies, and harvests ripe dates autonomously, ensuring minimal damage and improved efficiency. By integrating YOLO for object detection and CNN for maturity classification, the system optimizes harvesting decisions. This solution enhances productivity, reduces costs, and improves date quality, contributing to the advancement of agricultural automation.",Computer Science,2024 – 2025,"Robot, CNN, Object Detection, Date Harvesting",19
+P016,PulseXPert-Health Disorder Detection,"The aim of this graduation project, PulseXPert, is to develop an advanced smartwatch system for blood pressure management and health monitoring. The project addresses the need for continuous health tracking by providing real-time alerts, disorder detection, and machine learning-powered personalized insights and for accuracy. PulseXPert offers accurate health data without relying on external devices, enhancing initiative-taking care for both patients and healthcare professionals.",Computer Science,2024 – 2025,"Smartwatch, Technology Integrated System, Machine Learning in Health care.",20
+P017,Pandemic Alert System,"The ""Pandemic Alert"" project is a web-based platform that combines crisis management, artificial intelligence, and health informatics to improve epidemic preparedness and response. The system, which was created specifically for Saudi Arabia, helps the public, government health authorities, and hospitals to efficiently manage, track, and react to new health emergencies. The platform consists of three main modules: the Ministry of Health Module, which reviews cases, manages public health news, the Public Module, which provides citizens with real-time health alerts; and the Hospital Module, which logs patient cases, issues emergency alerts, and uses AI-driven data analysis. ""Pandemic Alert"" gives authorities the ability to take preventative action by offering an early warning system for epidemic tendencies, utilizing PHP and MySQL for secure data handling and AI algorithms for predictive analytics. By encouraging digital change in healthcare and working to create a robust infrastructure that can react quickly to public health emergencies, this project supports Saudi Arabia's Vision 2030.",Computer Science,2024 – 2025,"Artificial Intelligence, Pandemic, AI-driven data analysis",21
+P018,Wujud | وجود,"The way we interact, connect, and share knowledge has changed completely in today's constantly evolving digital environment. By developing an immersive platform that not only visualizes conferences but also improves attendee engagement, Wujud hopes to completely transform the conference experience. Wujud bridges the gap between in-person and virtual involvement with features including interactive displays, real-time chat with booth administrators, and customized agendas. Our platform guarantees attendees can participate completely in networking opportunities and content regardless of their physical location by utilizing advanced technology. Wujud, in our perspective, will not only boost conference attendance but also improve the whole event, making it more approachable and significant for all participants. We believe Wujud will not only boost conference attendance nevertheless also improve the overall experience, making it easily accessible and beneficial for everyone involved.",Computer Science,2024 – 2025,"Virtual Reality (VR), Augmented Reality (AR), Extended Reality (XR), Unified Modeling Language (UML), Corona Virus Disease (COVID-19)",22
+P019,صائب Drone detection in Dual Vision Modalities Using YOLO Models and SAHI Framework,"In line with Saudi Arabia’s strategic goal to localize 50% of its total government defense spending and achieve military self-sufficiency by 2030, this project delivers a cutting-edge solution to one of today’s most pressing and rapidly evolving security challenges: drone detection. We introduce a real-time drone monitoring system that fuses infrared (IR) and RGB imaging with advanced YOLOv11 object detection models, further optimized through the Slicing Aided Hyper Inference (SAHI) framework for improved small-object recognition. The system is engineered to detect small, fast-moving, and distant drones operating under complex and unpredictable real-world conditions, including variable lighting and background noise. This is achieved through a dual-model architecture that strategically balances inference speed, detection precision, and operational reliability. With scalability and adaptability at its core, the proposed solution significantly enhances situational awareness, early threat identification, and rapid response capabilities. It also contributes meaningfully to the expanding ecosystem of sovereign, locally developed defense technologies.",Computer Science,2024 – 2025,"Drone Detection, SAHI Framework, YOLOV11",23
+P020,Portfolio-Hub Faculty Portfolio Management System,"Today, faculty portfolios are considered an indispensable tool for faculty members across academic institutions. They serve as dynamic repositories of professional accomplishments, encompassing academic qualifications, teaching experience, scientific research, and a wide range of skills. In response to this need, PortfolioHub has been proposed as a centralized platform designed specifically for faculty members to document their achievements and showcase their expertise to peers, administrators, and students. The fundamental objective of this project is to implement a comprehensive digital portfolio system that enables faculty members to efficiently manage and update their professional information, including academic credentials, teaching history, research contributions, publications, awards, and other relevant activities. The proposed system will not only streamline the creation and management of portfolios but also enhance accessibility, organization, and visibility. By providing an intuitive and user-friendly interface, PortfolioHub will allow faculty members to build well-structured portfolios, easily associate supporting evidence, and maintain an up-to-date, professional digital presence that reflects their academic journey.",Computer Science,2024 – 2025,"Faculty portfolio, Digital Repository, Achievements, faculty members",24
+P021,Taifuna: Supporting Children With ASD,"Children with Autism Spectrum Disorder (ASD) often face considerable challenges in emotional expression, communication, and social engagement, which can significantly hinder their development and interaction with the world around them. To address these challenges, this project proposes the design and the implementation of a mobile application named Taifuna, specifically tailored to support children with ASD and their families. The application combines advanced drawing analysis tools with a wide range of educational activities designed to foster skill development, emotional understanding By analyzing children’s drawings using machine learning algorithms, the application provides insights into their emotional states empowering parents to better understand and support their children. Additionally, the application offers interactive, engaging learning activities to help children develop essential skills, including communication, cognitive abilities, and social interaction. This innovative solution aims to bridge gaps in emotional and educational support, ultimately improving the quality of life for children with ASD, strengthening family bonds, and fostering greater inclusion within the broader community",Computer Science,2024 – 2025,"Autism Spectrum Disorder (ASD), Drawing Analysis, Emotional Understanding, Machine Learning, Convolutional Neural Networks (CNNs).",25
+P022,Impact of Time Constraint on Maintenance Performance KPI,"The Vice Rectorate for Assets and Investment at Princess Nourah Bint Abdulrahman University is responsible for delivering maintenance and support services across various university facilities, in alignment with the institution’s broader goals. However, due to time-based and operational constraints, such as restricted access to campus areas during daily hours, crew performance is often misjudged using traditional Key Performance Indicators (KPIs) that fail to account for these limitations. As a result, maintenance tickets may experience unavoidable delays, particularly those with lower priority, negatively impacting overall performance scores. This project investigates the relationship between time constraints and maintenance KPIs through structured interviews and research. It proposes a system that adapts performance measurements to reflect real-world conditions more accurately. The developed solution consists of three main components: a machine learning model that classifies tickets and assigns difficulty-based weights, a dynamic KPI calculation module that adjusts based on those weights, and an interactive dashboard that visualizes performance across multiple styles. By integrating AI-driven predictions with flexible KPI design, the system ensures more accurate, fair, and actionable evaluations of maintenance performance under constrained conditions.",Computer Science,2024 – 2025,"Key Performance Indicators (KPIs), Time Constraints, Predictive Modeling, Machine Learning, Dashboard Visualization",26
+P023,Cloudy Robot,"Cloudy Robot is an interactive system that merges robotics with early childhood education, aimed at enhancing the learning and emotional development experience for young children. This project introduces a storytelling robot that communicates with children through engaging stories, providing comfort during moments of distress and facilitating vocabulary expansion. Recognizing that many parents face time constraints and challenges in finding interactive educational tools at home, Cloudy Robot addresses this gap by creating a unique solution. Additionally, The Cloudy Application will offer parental control and customization over the robot, allowing the parent to easily customize and schedule story playback, monitor their child’s interaction, and receive real-time alerts, such as notifications when the child is crying. Integrated face and sound detection features allow the Cloudy Robot to provide a safe, and personalized experience, blending learning, comfort, joy.",Computer Science,2024 – 2025,"Interactive storytelling, Robot, Facial recognition, Cry detection",27
+P024,Crime analysis and prediction,"This project proposes a predictive crime analysis system titled ""انتملاس"" Salamatna aimed at addressing rising crime rates through data-driven insights. By using machine learning algorithms, specifically the Random Forest Classifier, this system analyses past reported crimes to predict future incidents in different locations. Our project will provide a new aspect that make this system usable in both, a friendly way to connect our community, and help law enforcement, with critical insights to enhance public safety, effectively and most efficiently.",Computer Science,2024 – 2025,"Salamatna, Crime Analysis, Predictive Policing, Machine Learning, Random Forest Classifier, Community Safety, Friendly Neighbourhood Connections",28
+P025,Early Detection and Monitoring of Parkinson’s Disease Using a Mobile Application,"Parkinson’s disease (PD) is a progressive neurodegenerative disorder that significantly affects motor functions and quality of life. This project introduces an innovative iOS application designed for early detection and continuous monitoring of Parkinson’s symptoms using machine learning and data from wearable devices. The application features a user-friendly interface that allows patients to self-assess their condition, promoting remote healthcare and supporting clinical research. By leveraging advanced technology, this solution aims to enhance the accuracy of diagnosis and improve patient engagement, ultimately contributing to better management of Parkinson’s disease.",Computer Science,2024 – 2025,"Parkinson’s disease, machine learning, mobile application, early detection.",29
+P026,Potato Guard,"This project aims to address the problem of potato leaf diseases such as early blight and late blight and reduce the resulting losses and their impact on global crop productivity. The application is designed using MobileNetV2 CNN to detect the disease by uploading the image during the application, where it is processed, and a suitable treatment plan is suggested. The application also allows expert consultation. The main expected output is an accurate system called ""Potato Guard"".",Computer Science,2024 – 2025,"Deep learning, Convolutional Neural Network, Late blight, Early blight, MobileNetV2 CNN",30
+P027,ParkPro,"ParkPro is about offering easy way reserving parking spots, solving congestion, reducing searching time for a parking space, which is the main issues that face university students and employees in Saudi Arabia. ParkPro's system is designed with the Agile methodology in mind for iterative development and continuous user feedback. It lets users search, reserve, and rate parking spots, and the key features are marking parked car location and real-time notifications when spots become available",Computer Science,2024 – 2025,"Internet of Things, Global Positioning System, Application Programming Interface, Local Area Network, Wide Area Network",31
+P028,Easy Picks,"In a world increasingly reliant on online shopping, the need for applications that provide a seamless and convenient experience for users has become essential. This is where Easy Picks comes in—the app that combines advanced technology with an easy-to-use interface to offer you an outstanding shopping experience. You can easily search for products either by entering their names or by uploading an image of the desired item, as the app analyzes the image to provide accurate options. The app also includes a dedicated section for saving printed invoices and converting them into electronic format, making it easy to track your purchases and use these invoices in the recommendation system. The recommendation system is based on your preferences, ensuring that you receive suggestions tailored to your interests and needs. We believe in the importance of listening to our customers, which is why the app allows you to share your personal experiences and submit ratings, contributing to service improvement. Additionally, making it easier to find the perfect options within your budget. The Easy Picks app is not just a shopping tool; it is a comprehensive platform designed to provide a comfortable and efficient user experience.",Computer Science,2024 – 2025,"scanning, image search, online shopping, Recommendation system, artificial intelligence.",32
+P029,Collab: A Revolutionary web-based platform designed for collaborative coding projects,"Collab is a web-based platform developed to transform collaborative coding by addressing key challenges such as poor communication, version control issues, and fragmented workflows. It provides two core features that enhance productivity and teamwork. The first is real-time simultaneous coding, which allows multiple users to write, review, and edit code together in a shared environment. The second is integrated project management, offering drag-and-drop UML diagramming, task distribution, and progress tracking tools—all within the same workspace. Designed especially for academic and educational settings, Collab streamlines the collaborative process by combining technical development with visual project planning. Its intuitive interface supports various programming languages and tools, ensuring accessibility for students, instructors, and development teams alike. With full browser-based access and support for remote collaboration, Collab is built for modern, global use. It empowers users to manage projects efficiently, foster innovation, and improve outcomes in coding-based coursework, capstone projects, and real-world software development practice.",Computer Science,2024 – 2025,"Collaborative Coding, Real-time Collaboration, Project Management Tools, Version Control, Educational Platform",33
+P030,FunColor,"The concept of playing and learning in the classroom has begun to change from playing using physical toys to playing using a computer, such as games. Which now finds its counterpart in games on the computer. Games are probably the most common method of learning today, especially for younger ones because they serve as an interactive open door from which kids access information during play activities. Thus, in this case, the project is aimed at providing an overall engaging way of teaching children colors by merging play and learning. Different activities such as experiments and sensory experience have been combined into a mobile application that contains simple color games and interactive tools to form a real deep understanding of Color Theory and its applications in the Real World. All this will enable children to provide an understanding of how colors will in various circumstances form the basis for educational, critical thinking, creativity, and problem-solving activities. The education of children about colors through playing educational games will not only serve to introduce them to practical end-options like color definitions, but it would also nurture their mind development. This ensures learning experience will be fun and engaging while improving children's ability to grasp concepts in a memorable way. .",Computer Science,2024 – 2025,"Color, Games, Application, Education, Cognitive Development.",34
+P031,بصيرة Basirah,"Human Resources is essential for finding and hiring the right people, helping organizations grow and succeed. However, traditional resume screening methods cannot keep up with the needs of today’s competitive job market. Manual processes are often slow, resource-intensive, and can be unfair due to unconscious biases. This leads to slow hiring, less workplace diversity, and missed chances to hire the best candidates. The main challenge lies in managing large volumes of applications efficiently and fairly. AI-powered tools offer a solution by streamlining the recruitment process, reducing biases, and improving hiring outcomes. By automating resume screening and leveraging advanced algorithms, organizations can evaluate candidates faster, more accurately, and more objectively. Baseera specifically addresses these challenges by providing an AI-powered system for automating resume screening. By applying Natural Language Processing (NLP) techniques, such as Named Entity Recognition (NER) and GPT-enhanced models, Baseera accurately extracts key information from resumes, including skills, education, and experience. These AI algorithms help reduce human bias, streamline candidate evaluations, and improve the overall quality of hiring decisions. Through automation, Baseera significantly reduces the time and resources needed for initial candidate screening while enhancing fairness and objectivity. By integrating advanced AI technologies, organizations can achieve faster, smarter hiring decisions and build more diverse, high-performing teams..",Computer Science,2024 – 2025,"NLP (Natural Language Processing), NER (Named Entity Recognition), AI (Artificial Intelligence), HR (Human Resources)",35
+P032,دُرَّة DURRA – Ovarian Cancer Diagnosis Using Machine Learning,"Ovarian cancer remains a leading cause of cancer-related deaths among women, primarily due to the difficulty of detecting it in early stages. Symptoms are often vague and non-specific, leading to delayed diagnosis and significantly lower survival rates. Traditional diagnostic methods—such as CA125 biomarker testing, imaging (ultrasound, CT, MRI), and biopsy—are limited by low sensitivity, variability, and time constraints. Recent advancements in Artificial Intelligence (AI), particularly machine learning, offer a promising alternative by enabling the analysis of complex clinical data with high precision and consistency. This project presents Durra, an AI-based diagnostic web system designed to support early detection of ovarian cancer. Durra utilizes the XGBoost algorithm, known for its superior performance with structured medical data, to build a predictive model using clinical features including CA125, HE4, CEA, NEU, age, and menopause status. The model was trained and evaluated using two real-world datasets—Mendeley and a publicly available dataset from GitHub. It achieved test set accuracies of 97.14% and 95.71%, respectively, demonstrating high reliability and generalizability. Durra not only improves detection accuracy but also provides rapid predictions through an integrated web platform accessible to doctors. By enabling earlier diagnosis, reducing false positives, and supporting more informed treatment decisions, Durra highlights the impactful role of AI in advancing women’s health and improving cancer care outcomes.",Computer Science,2024 – 2025,"Machine Learning (ML), Extreme Gradient Boosting (XGBoost), Ovarian Cancer (OC)",36
+P033,راصد RASID | Road defect detection and classification,"The maintenance of road infrastructure is important for ensuring safety and efficiency in transportation systems. Current manual inspection methods for detecting road defects, such as potholes, cracks, and manholes, are inefficient and lead to delayed responses . This project proposes RASID, a smartphone-based road defect detection and classification system specifically designed for Riyadh. Utilizing advanced technologies such as Convolutional Neural Networks (CNNs) and the YOLOv8 algorithm, the system automatically classifies road defects with high accuracy. The application integrates the Google Maps API to capture and report precise defect locations, enabling users to submit reports directly to the Maintenance team. By streamlining defect detection and reporting, RASID aims to enhance road safety, reduce accidents, reduce driver stress, and improve the overall maintenance process within the city of Riyadh.",Computer Science,2024 – 2025,"CNN, Deep learning,Road Defect, Detection and Classification.",37
+P034,Visualize,"Visually impaired people have trouble navigating around, identifying objects, reading and translating texts, and accessing information. Current assistive technologies often cover different needs independently, so that users often rely on multiples of these tools in their daily lives, which may be inefficient and impossibly inconvenient. To solve this problem, we developed ""Visualize"", a mobile integrated application which leverages the recent advancements in Artificial Intelligence, Computer Vision, Natural Language Processing (NLP). Visualize brings together into one user-friendly app the most important features: navigation, object recognition, reading text aloud, text translation, and searching for information. Completely voice-activated, the app allows an intuitive user interface designed for the unique needs of visually impaired users. It features leading technologies like OCR in extracting text, AI-powered real-time object detection, and straightforward voice interaction. The application is designed to help users' communication with the outside environment, giving them more independence and confidence in everyday tasks. Offering one solution, Visualize is designed to fix this problem in existing assistive tools and offers an integrated unified tool, that is considered a practical solution that would add real value to the quality of life for people with impaired vision. This project represents a meaningful leap forward in assistive technologies.",Artificial Intelligence,2024 – 2025,"Visual Impairment, Navigation, Object Recognition, Text Reading, Voice Commands.",39
+P035,دال AI behavioural analysis for academic decision making (DALL),"Choosing the right college major is a critical decision for high school students, as it significantly influences their academic and professional future. However, many students struggle with this decision due to a lack of personalized guidance and an overwhelming number of options. This project proposes a solution to address these challenges by developing a Major Guidance Test system that combines GPA-based eligibility analysis and interest-matching to provide tailored recommendations. The system incorporates key features such as a GPA calculator, which computes the student’s academic standing based on school grades and standardized test scores. This GPA is then used as an essential input for the model, ensuring that the recommendations are both realistic and achievable. Additionally, the platform includes an interactive chatbot to assist students by answering questions about majors, eligibility requirements, and career paths. By integrating these tools, the system aims to empower students with personalized recommendations that align with their strengths, preferences, and academic qualifications. This solution not only reduces uncertainty but also helps students confidently choose a major that sets them on a successful academic and career path. The platform has the potential to transform how students approach one of the most important decisions in their educational journey.",Artificial Intelligence,2024 – 2025,"Major Guidance Test, Personalized Recommendations, Academic Decision-Making.",40
+P036,رَفد RafdSign: Audio-to-Sign Language Translator,"Individuals with hearing impairments face significant challenges in communicating with others in their daily lives, even during simple interactions. These communication barriers can lead to social isolation and limit opportunities for education, employment, and everyday interactions. Our project aims to address this issue by developing a speech-to-sign language translation system. This system will convert spoken language into sign language, helping bridge the communication gap between deaf and hearing individuals. Currently, solutions that effectively handle sign language translation are limited due to the complexity of gestures, expressions, and context in sign languages, posing a significant barrier for those who rely on it for communication. By addressing this need, our project aims to provide a tool for individuals with hearing impairments, making everyday communication more accessible for everyone. Our solution leverages AI models to ensure accurate translation. The model first converts the audio signal to text. Then, the text is converted to gloss, and finally, the gloss is used to extract keypoints that are utilized to render the final video. Our system is tested against multiple state-of-the-art models using various evaluation metrics and across several datasets. Additionally, we will evaluate our model through a user study to capture aspects that cannot be assessed using quantitative evaluation metrics.",Artificial Intelligence,2024 – 2025,"Sign Language Translation, Speech-to-Sign Language, Assistive Technology, Hearing impairments, Generative Adversarial Networks (GANs)",41
+P037,Children’s Educational Application Battoot,"This project introduces Battoot, an AI-powered educational application designed to support early Arabic literacy among children aged 4 to 8. The application aims to bridge the gap in digital Arabic learning tools by focusing on two essential skills: handwriting and pronunciation. Battoot integrates a fine-tuned Convolutional Neural Network (CNN) for Arabic Handwriting Recognition (HWR) and a pre-trained transformer-based model for Automatic Speech Recognition (ASR). The system provides real-time, constructive feedback to learners, enhancing engagement through gamified elements and a child-friendly interface. A robust client-server architecture ensures efficient processing, while model performance is validated through extensive fine-tuning and testing on child-specific datasets such as Hijja and Dhad. The result is an innovative educational application that not only improves children’s recognition and production of Arabic letters but also sets a foundation for further language development. By combining deep learning, thoughtful design, and interactive learning principles, Battoot contributes a novel solution to early childhood Arabic education.",Artificial Intelligence,2024 – 2025,"S Arabic Handwriting Recognition, Speech Recognition, Gamification, AI-fine-tuning models",42
+P038,UniFix— Predictive Maintenance,"Predictive maintenance is a modern approach to facility management that leverages artificial intelligence to anticipate equipment failures before they occur. In this project, we developed UniFix—an intelligent predictive maintenance component designed to be integrated with the University Agency for Facilities and Operations at Princess Nourah bint Abdulrahman University. The system applies unsupervised machine learning algorithms to historical maintenance records, enabling it to detect anomalous patterns and cluster assets based on failure behaviors. Without relying on sensor data or pre-labeled events, UniFix extracts insights directly from real operational records and generates interpretable predictions about the expected failure date of each asset. By identifying behavioral similarities among assets, the system enhances awareness of maintenance-related risks and supports proactive planning. UniFix was built with flexibility and scalability in mind, allowing it to evolve and integrate into broader maintenance platforms. This project demonstrates how artificial intelligence, when applied thoughtfully, can transform traditional maintenance practices into predictive, data-driven solutions that improve operational efficiency and support strategic decision-making.",Artificial Intelligence,2024 – 2025,"Failure Cycles, Unsupervised Learning, Clustering, Resources Optimization.",43
+P039,MUKHBIR: A Hybrid Framework for Detecting GAN-Generated Images,"With AI-generated images becoming increasingly realistic, the challenge of detecting synthetic content is more critical than ever. Traditional detection methods often struggle to generalize across different generative models or lack transparency in how decisions are made. Mukhbir is a hybrid detection system designed to address this gap by combining gradient-based image transformation with a SwinV2-Large Transformer classifier. Tested on a balanced and high-quality dataset from seven different GANs, Mukhbir offers both strong accuracy and visual explainability, supporting real-world use in media verification and digital forensics.",Artificial Intelligence,2024 – 2025,"AI Generated Images, Generative Adversarial Networks (GAN), Binary Classification, AI Ethics, Digital Media Integrity.",44
+P040,LADON,"Children who experience stuttering face significant challenges in communication, often impacting their self-esteem and social interactions. Early intervention and engaging methods of therapy are crucial in supporting these children to overcome their stuttering and improve their quality of life. We introduce our project LADON, an innovative, gamified application designed to support children with stuttering through personalized therapy. LADON application features a parent sign-up process, automated voice analysis to identify stuttering patterns, and interactive games tailored to address individual needs. Cartoon characters serve as virtual companions, guiding children through their therapy journey in a fun and engaging way. Additionally, the application offers progress tracking and detailed feedback for parents, ensuring a collaborative approach to therapy. By incorporating gamification, LADON transforms traditional speech therapy into an enjoyable experience while fostering motivation and consistency. As children progress, the platform adapts by increasing game difficulty to match their improvement. LADON will be built using Swift, which will combines cutting-edge technology with a user-centered design to make therapy more accessible and effective. Our project highlights the potential of gamified environments to enhance therapy outcomes, offering a meaningful solution for children who stutter and their families. By that we aim to bridge the gap between therapy and engagement, empowering children to achieve better communication skills and greater confidence.",Artificial Intelligence,2024 – 2025,"Stuttering, Child Speech Therapy, Gamification, Voice Recognition, Adaptive Learning, Personalized Therapy.",45
+P041,PNU ChatBot,"The “PNU ChatBot” project aims to help students easily access academic information through a web-based chatbot. Using natural language processing (NLP), it provides 24/7 instant responses to common questions about programs and services, improving student engagement and self reliance. The chatbot is developed through stakeholder input and undergoes design, testing, and deployment on the web, with feedback gathered for continuous improvement.",Artificial Intelligence,2024 – 2025,"Higher Education Chatbot, Student Information System, Campus Resources Automation, Real-Time Information Retrieval, Academic and Administrative Support, AI-Powered Student Services, User-Centric Design",46
+P042,Luminis,"Cognitive systems are now essential tools that help in constructing and enhancing mind performances, in which people can usually apply in active intellect, critical thinking, and problem-solving. These systems have evolved from cognitive science and have moved from being static to becoming more responsive and adaptive. On this basis, Luminis is a web application that offers an intelligent and engaging approach for users to study and practice their intellectual skills by solving numerous and various different text-based puzzles. It has a clear and easy-to-use interface that offers feedback and tracks the user’s progress in order to have a constant and measurable development. Luminis provides a practical solution to the problems of the traditional cognitive training tools such as non-personalized approach, repetitive inputs, and lack of engagement, making cognitive development more available, effective, and fun. In the long run, Luminis wants to bring about long-term cognitive development through the use of interactive, goal-oriented, and feedback-based learning methods that are fully personalized to the user’s unique needs and evolving performance.",Artificial Intelligence,2024 – 2025,"Luminis, Artificial Intelligence, Cognitive Training, Natural Language Processing.",47
+P043,Alfaba - ألفباء,"Alfaba is an educational game tailored for children with dyslexia, designed to support Arabic language learning through interactive storytelling and structured assessments. The game is organized into three levels (gates), each focusing on a specific positional form of Arabic letters: initial, medial, and final. Stories are dynamically generated using ALLaM, an Arabic language model developed by SDAIA and fine-tuned on a custom dataset to ensure age-appropriate and engaging content. The model was evaluated using both automatic and human evaluation methods, demonstrating strong performance in coherence, fluency, and adherence to instructions. To support diverse learning styles, the game incorporates educational voice narration that aids auditory processing. Children’s progress is monitored through periodic assessments following each story and a comprehensive final test, evaluating their mastery across all levels. The game is developed using Godot, an open-source game engine, along with Visual Studio Code and Adobe Illustrator for interface design. This project aligns with Saudi Vision 2030 by fostering educational innovation and inclusivity, enhancing literacy among children with learning difficulties, and contributing to the digital transformation of the education sector through the integration of advanced AI technologies.",Artificial Intelligence,2024 – 2025,"Serious games, Generative Artificial Intelligence, Learning Disabilities, Dyslexia, Storytelling, Arabic language, Large Language Models, ALLaM.",48
+P044,Nabeeh : Artificial Intelligence-Based System for Attention monitoring of students,"A This project presents an AI-powered system designed to monitor and enhance student engage ment in classrooms through automated video analysis. By processing uploaded classroom recordings, the system detects and analyzes each student’s face frame-by-frame using two key components: Ef ficientNet, fine-tuned on two child-focused datasets for emotion recognition with an accuracy of 84%, and OpenFace for gaze tracking. The results of these components are then aggregated through the trained machine learning model SVM with an accuracy of 96% to predict the engagement of the frame using the outputs of both of the EfficientNet and OpenFace. The extracted engagement data is aggregated and visualized through insightful charts -includ ing pie charts for overall engagement distribution and smoothed line graphs for time- based attention trends. These visualizations are compiled into a downloadable PDF report, enabling teachers to re view class performance at a glance and identify when students tend to lose focus. The system also stores engagement summaries in a real-time database, allowing teachers and administrators to track progress across multiple sessions. By providing objective, data-driven feed back, this solution addresses major limitations of manual observation, such as subjectivity and lack of scalability. Designed with privacy, usability, and educational value in mind, this AI system bridges the gap between traditional classroom monitoring and modern, data-informed teaching strategies, making it especially effective for engaging younger learners.",Artificial Intelligence,2024 – 2025,"Artificial Intelligence (AI), Attention Monitoring, Gaze Tracking, Emotion Recognition",49
+P045,Thaqib | Enhancing Critical Thinking with Multimodal Generative AI,"The integration of Artificial Intelligence (AI) in education has significantly transformed traditional learning experiences. However, increased reliance on generative AI technologies for immediate solutions, particularly in mathematics, has raised serious concerns regarding the erosion of students’ critical thinking skills. Students frequently depend on AI-generated answers, bypassing essential cognitive processes such as independent reasoning and problem-solving, ultimately hindering creativity and autonomous thought. Thaqib is an innovative AI-powered educational platform designed to enhance critical thinking and cognitive reasoning skills in mathematics among sixth-grade Saudi students. By integrating storytelling and Socratic questioning, Thaqib actively engages students, facilitating deeper understanding and problem-solving capabilities. The platform utilizes advanced Large Language Models (LLMs) to dynamically generate mathematical narratives tailored to students’ learning levels. Additionally, Thaqib leverages AI-driven image generation to create contextual visual representations, enhancing comprehension and engagement. External knowledge retrieval techniques are employed to ensure content accuracy, reduce errors, and support adaptive, personalized learning experiences for each student.",Artificial Intelligence,2024 – 2025,"Generative AI, Large Language Models (LLMs), Image Generation, Retrieval Augmented Generation, Critical Thinking",50
+P046,SaudiWay,"SaudiWay is an application designed to enhance the tourism experience in Saudi Arabia. Along with having a 24/7 Al tour guide, personalized suggestions for lists, and in-app booking of activities for easing the tourists journey, it caters to both local and international tourists; the interface is in English and the Al tour guide translates into various languages. Moreover, SaudiWay is not only a tool for planning trips but also supports local businesses by allowing them to share their activities on the app which will help in boosting the country's economy. This focus on local collaboration enables tourists to experience the real culture of Saudi Arabia, therefore providing more immersive travel experiences.",information systems,2024 – 2025,"Saudi Tourism, Artificial Intelligence, Smart Tourism, Personalized Travel, Multilingual Tour Guide, Local Businesses, Cultural Tourism, In-App Booking",52
+P047,ASMAAK | أسمعك,"The “ASMAAK | أسمعك” project is a child-friendly educational mobile application designed to support children with hearing and speech impairments through the learning of Arabic Sign Language (ArSL). The app aims to bridge the gap in accessible educational tools in the Arabic-speaking world by offering a visually engaging, interactive, and culturally appropriate experience tailored to children’s cognitive abilities and learning styles. The application features intuitive navigation, colorful visuals, and gamified elements to enhance user engagement and motivation. Through playful learning scenarios, children are encouraged to recognize and practice basic signs, building both confidence and communication skills. The design process followed a user-centered approach with continuous feedback from target users and specialists in speech therapy and accessibility. The application also emphasizes inclusivity and aims to foster a sense of belonging and independence for children with disabilities. “ASMAAK” aligns with the Saudi Vision 2030 goals of enhancing digital solutions and empowering people with disabilities by offering an innovative, tech-based approach to inclusive education.",information systems,2024 – 2025,"Arabic Sign Language, Accessibility, Inclusive Design, Educational App, Child UX",53
+P048,Mustadam | مُستدام,"This project proposes a platform called ""Mustadam"", a mobile application designed to promote sustainability on university campuses by encouraging recycling and reusability among students and staff. The app utilizes AI-powered image analysis to assist users in correctly sorting recyclable materials, helping reduce waste, pollution, and environmental harm. One of its key features is the Mustadam Hub, a system that facilitates the donation and borrowing of books, school supplies, and electronics. This feature is supported by a point-based rewards system that motivates users to participate actively. In addition, the app includes a Recycling History section, allowing users to track their individual contributions over time. It also provides information on different recycling symbols and methods and offers a map showing the nearest recycling bin locations. A university-wide leaderboard highlights the most environmentally active departments, encouraging teamwork and friendly competition. Mustadam merges innovation, environmental education, and technology to build a stronger culture of sustainability at Princess Nourah bint Abdulrahman University.",information systems,2024 – 2025,"Princess Nourah University (PNU), Artificial Intelligence (AI), Kingdom of Saudi Arabia (KSA), Database (DB), Sustainability.",54
+P049,ثرا Thara Application,"Digital platforms can be utilized in the livestock sector to achieve sustainability, economic stability, and improved quality of life by facilitating access to national products, enhancing market efficiency, and enhancing the capabilities of Livestock Producing Families (LPF). The method of development of the mobile application follows the iterative waterfall method. This simple System Development Life Cycle (SDLC) approach allows for continuous feedback, real-time adjustments, and system improvements. The 'Thara' application is designed to support the livestock market by providing training programs from the Ministry of Environment, Water, and Agriculture (MEWA), offering livestock tools from MEWA or various suppliers, selling livestock products, and consulting with the MEWA Adviser. The ""Thara"" application has been designed to support the LPF by providing a user-friendly interface, relevant training programs, and a digital environment for the livestock market. The ""Thara"" application's goal is oriented toward enhancing sustainability, economic growth, and the quality of life for LPF members. This digital tool ensures more efficient access to resources, fosters innovation, and promotes long-term growth for livestock families.",information systems,2024 – 2025,"LPF, livestock, MEWA, MEWA Adviser, training programs.",55
+P050,بُؤرة Bu’rra: A Real-Time Infant Safety Application for Choking Detection,"The Bu’rra application is an innovative safety solution designed to enhance the well-being of infants by monitoring their facial expressions, body movements, and skin color to detect potential choking incidents. Using real-time analysis, Bu’rra leverages an external webcam to continuously observe and assess the baby’s physiological indicators, such as facial distress and changes in color, as well as physical movements. The system is engineered to provide instant alerts to parents, caretakers, and healthcare professionals, ensuring rapid response in critical situations. This application is tailored for children aged zero to two years, a vulnerable age group that requires constant supervision. By focusing on early detection, Bu’rra offers a proactive approach to preventing life-threatening choking episodes. The intuitive interface allows parents and caregivers to monitor the baby’s safety effortlessly, while healthcare professionals can also use the system for enhanced care in clinical settings. Bu’rra’s real-time monitoring capabilities are designed to ensure timely intervention when every second matters. The alerts generated by the system help minimize delays, allowing immediate action to be taken by those responsible for the infant’s care. This innovative tool is set to transform infant safety, providing peace of mind for parents and caretakers while supporting healthcare assistants in delivering optimal care for vulnerable children. By incorporating cutting-edge technology into everyday caregiving, Bu’rra is poised to become an essential part of infant care and safety practices.",information systems,2024 – 2025,"Infant Safety, Choking detection, AI-Powered Safety, Machine learning",56
+P051,Sappaq - سبّاق,"The “Sappaq” project is an innovative technological solution developed to address the growing issue of traffic congestion in the Kingdom of Saudi Arabia, particularly congestion caused by road accidents. Recognizing the delays and inefficiencies in handling traffic incidents, Sappaq leverages drone technology to provide rapid and accurate accident documentation. The system is integrated with Google Maps and works in coordination with the Najm accident reporting service. When an accident is reported through Najm, a drone is automatically dispatched to the scene using pre-programmed GPS coordinates. Upon arrival, the drone captures high-resolution images and video footage of the accident site from multiple angles. This real-time visual data is then transmitted to a centralized control center, where specialists can quickly assess the situation, determine responsibility, and begin the necessary procedures without waiting for manual on-site inspections. This approach not only speeds up the accident documentation process but also reduces the impact on traffic flow by minimizing the time vehicles spend at the scene. Furthermore, the use of drones enhances the accuracy of accident reports and ensures safer, contactless data collection. “Sappaq” is a prime example of how smart technology and automation can be harnessed to improve traffic management, emergency response, and overall urban mobility in modern cities.",information systems,2024 – 2025,"Sappaq, Drone Technology, Accident Documentation",57
+P052,نبر Nabrah,"Mental health disorders, including depression and anxiety, are among the most prevalent global health concerns, affecting millions of individuals. Early detection and diagnosis remain key challenges due to the reliance on self-reported symptoms and the stigma surrounding mental health issues. Nabrah application aims to bridge this gap by utilizing voice biometrics to provide an innovative, and accessible method for mental health assessment. The application leverages artificial intelligence (AI) and machine learning (ML) algorithms to analyze speech patterns, tone, pitch, and rhythm, extracting emotional andpsychological indicators. By detecting subtle changes in voice, Nabrah offers users insights into their mental well-being, allowing them to track emotional changes over time and seek professional help when needed. The system is designed to be user-friendly, private, and secure, ensuring that users can record their voices and receive feedback in a safe environment. Nabrah integrates advanced speech analysis models trained on diverse datasets to recognize emotional states effectively. Unlike traditional methods that require in-person consultations, this technology provides real-time assessments, making mental health support more accessible and immediate. The application is also beneficial for healthcare professionals, as it can aid in preliminary screening and continuous monitoring of patients. A comparative analysis with existing applications highlights Nabrah’s strengths in voice-based AI diagnosis, real-time emotional analysis, and personalized recommendations. Security is a primary concern, and the app ensures data encryption and compliance with privacy regulations to safeguard user confidentiality. While challenges such as cultural adaptation and model accuracy exist, Nabrah represents a significant step forward in integrating technology with mental health care. By providing a scalable, AI-driven solution, this project has the potential to revolutionize early mental health diagnosis and improve well-being across diverse populations.",information systems,2024 – 2025,"Mental Health, Voice Biometrics, Machine Learning, Artificial Intelligence",58
+P053,AI Grading Application,"Manual grading of essay-style questions presents significant challenges for educators, including time consumption, inconsistency, and subjectivity. To address these issues, we developed GradeGenius, an AI-powered application that integrates ChatGPT to assist teachers in grading student responses. This system enables educators to upload exam content, supporting materials, and student information, after which it automates the grading process and provides feedback. GradeGenius streamlines the assessment process while maintaining accuracy, fairness, and efficiency. The solution is especially suited for handling large volumes of assessments and offers statistical analysis to support data-driven teaching strategies. While not a replacement for human judgment, GradeGenius enhances the grading experience by reducing workload and improving feedback quality in modern educational environments.",information systems,2024 – 2025,"AI grading, ChatGPT, essay evaluation, GradeGenius, automated assessment",59
+P054,Rahalah - رحاله,"Through the ""Rahalah"" app Saudi Arabian tourism will experience transformation because it solves essential problems that affect both tourists and the tourism sector namely cultural unfamiliarity and inadequate vacation planning. The application serves two important functions through its cultural education of Saudi Arabia and its smart trip planning functions and its connection with local student guides to support responsible tourism. The methodology supports cultural education and cooperative training programs for tourism students because they facilitate Saudi heritage site exploration while advancing the economic goals of Vision 2030 through tourism revival. A better enriched travel experience will result from this approach which will simultaneously close cultural differences and protect heritage sites while making Saudi Arabia an elite worldwide tourism destination that benefits both foreign visitors and residents.",information systems,2024 – 2025,"Rahalah app, tourism, training, cultural exploration, historical attractions, booking services, guide interactions.",60
+P055,Tabayn (تبیّن),"Smishing attacks continue to increase as a cybersecurity threat. There are many tools available to detect them, but the gap we want to fill is that most of them are not designed to consider the linguistic and cultural differences of non-English languages, leaving Arabic-speaking users more vulnerable. The need for smishing detection arises because such attacks cause significant financial and personal damage, while traditional tools typically fail to address the unique challenges of Arabic text messages. This project presents Tabayn, a mobile application developed to detect and prevent smishing in Arabic text messages. The importance of Tabayn lies in its ability to protect Arabic-speaking users from smishing attacks in their most dangerous forms. Tabayn leverages the latest Natural Language Processing (NLP) and Machine Learning (ML) techniques to provide accurate analysis and classification of text messages and identify malicious content. To enhance model performance, the project implemented the stacking method and ensemble learning. The base models used in the stacking method include Support Vector Machine (SVM) with a linear kernel, Random Forest (RF), and Gradient Boosting, creating a robust foundation for reliable detection. The meta model used for final classification is Logistic Regression. Additionally, a complete dataset of smishing text messages in Arabic was developed, collected through user surveys and enriched with AI-generated examples, ensuring linguistic authenticity and diversity. A waterfall methodology was employed, ensuring structured development from requirements gathering to system deployment. The goal of Tabayn is to provide users with a seamless way to upload messages for verification and receive immediate feedback, ultimately increasing their awareness of smishing tactics.",information systems,2024 – 2025,"Smishing detection, Arabic text, Machine learning (ML), Natural language processing (NLP), Stacking.",61
+P056,دليلة Delelah Application,"Imagine that your personalized trip plan in Saudi Arabia appears in one click! The Delelah application is designed to transform tourism planning in the Kingdom of Saudi Arabia. Considering the rapid growth in the tourism sector, tourists face multiple challenges, including the lack of platforms or applications that accurately meet their needs and preferences. The Delelah application addresses these challenges by providing personalized plans that cater to users' preferences, available attractions, and budget constraints. With its intuitive design and advanced features, Delilah guides Users through the Kingdom of Saudi Arabia's rich cultural and entertainment landscape and ensures a seamless and enjoyable travel experience. By integrating local events, top destinations, and seasonal highlights, Delelah aims to redefine how tourists explore and experience the wonders of Saudi Arabia.",information systems,2024 – 2025,"Saudi Arabia tourism, Personalized Plans, Tourist preferences",62
+P057,PopList,"This project aims to create a mobile application called PopList which helps ADHD individuals as they face difficulty in following up on their daily tasks and focusing on them. PopList allows the users to easily list their daily tasks so that they can complete them. The task will be in the form of a bubble and when the task is completed, the user will be able to click on the bubble and it will pop and disappear. The goal for the user is to complete all their daily tasks and when they are all completely finished, a goldfish character will appear and show the users a motivational message upon the tasks completion. Afterwards, the user will have access to a fun and useful maze game dedicated to ADHD individuals that helps to increase focus and reduce distraction. The reason for choosing the goldfish as the main character is due to its average focus span which is 9 seconds, while the focus span of a normal person is 8 seconds. PopList aims to raise the user's level of focus to a higher level than the focus span of a goldfish. It’s a fun and challenging maze game with different levels of complexity to suit ADHD individuals. The project follows the waterfall methodology, as the scope and requirements are specific and clear, and the requirements are derived and collected through surveys. The analysis and design of this project depends on UML diagrams.",information systems,2024 – 2025,"ADHD, focus span, distraction, goldfish.",63
+P058,نَبْت A Machine Learning-Based Plant Recommendation System for Urban Sustainability in Riyadh,"Urban greening has become increasingly important, especially in Saudi Arabia, where Vision 2030 emphasizes sustainability and environmental well-being. This project focuses on developing a machine-learning-based plant recommendation system tailored to Riyadh's unique climate, using machine learning models such as K-nearest neighbor (KNN) and K-means. The system analyzes inputs such as spread, humidity, irrigation, maintenance, plant type and salinity to recommend suitable plant types, supporting initiatives like the Green Riyadh project. By combining urban planning with advanced technologies, the system aims to assist decision-makers in selecting plants that thrive in specific conditions, thereby contributing to the creation of sustainable and livable urban spaces. The project addresses challenges like limited data and scalability. It offers a user-friendly dashboard for recommendations and plant tracking. This innovative approach aligns with Vision 2030’s goals and represents a step toward smarter urban planning.",information systems,2024 – 2025,"Vision 2030, Green Riyadh project, Agriculture, Machine learning (ML), Recommendation system.",64
+P059,Blood Bank Inventory System - Artificial Intelligence (AI) Powered Blood Bank Dashboard,"Healthcare institutions are usually challenged by the management of blood inventories due to the failure of traditional blood banks to accurately predict donation needs. This paper offers a remedy through the creation of a machine learning-driven intelligent blood bank inventory dashboard system to address such challenges effectively. The system adopts advanced algorithms, including Prophet, logistic regression, and random forests, to analyze historical data and project future trends in blood demand and utilization. With these predictive capabilities, the dashboard enables blood banks to optimize stock levels, significantly reducing wastage and providing sufficient blood supply during emergencies. The system also supports efficient inventory management with real-time insights into patterns of blood consumption and donation. This proactive capability not only enhances operational efficiency but also equates to improved patient outcomes by having life-saving blood supplies available when needed. The introduction of machine learning in blood inventory management represents a groundbreaking development within healthcare logistics. The implementation of this novel solution can lead to more reactive and streamlined blood bank operations, with the overall impact of fostering better resource planning and enhancing the level of care provided to patients. Through this system, we hope to revolutionize the blood banking system to bring about a new era in which accurate forecasting and optimum management are the standards of blood inventory management.",information systems,2024 – 2025,"Machine learning, inventory management, blood donation, demand forecasting, healthcare transformation.",65
+P060,Aoun - عون,"The Aoun app addresses this need by offering an integrated platform that connects car owners with service providers, enabling users to request towing services, rent vehicles, arrange repairs, and order spare parts with ease. Designed with a user-centric approach, Aoun ensures seamless communication between drivers, service providers, and other users, reducing service delays and minimizing traffic disruptions caused by car breakdowns. The project employs an iterative waterfall methodology, incorporating continuous feedback to enhance system functionality at each development stage. Beyond solving immediate roadside issues, Aoun contributes to broader goals by promoting local businesses, enhancing transportation quality, and supporting the economic diversification outlined in Saudi Arabia’s Vision 2030. By streamlining critical processes and fostering community collaboration, Aoun not only improves the on-road experience but also strengthens the infrastructure for sustainable growth and social development in the Kingdom",information systems,2024 – 2025,"Emergency assistance, car owner community, service provider integration, Aoun application.",66
+P061,Woman's Car Club,"The Women’s Car Club project addresses the growing need for a dedicated, supportive platform that empowers women drivers at Princess Nourah University in Saudi Arabia. As driving becomes increasingly accessible to women in the region, new challenges and opportunities arise, especially in a field that has traditionally been dominated by men. The application shall fill the gap by offering a collaborative platform that would assist and support women in making decisions on car ownership, buying and selling and expert advice. Aligned with Saudi Arabia’s Vision 2030, which promotes women’s empowerment across all sectors, the Women’s Car Club serves as a pioneering step towards enhancing women’s autonomy and confidence in automotive-related decisions. The advantages of the Women's Car Club application support artificial intelligence, a marketplace for buying and selling cars, a rating system, and communication with experts, as well as reviews of various types of cars. The system allows users to register, log in and log out as well as create, edit and delete their posts while allowing the ability to comment to other user’s post. Also, the system support search and notification function. The application boasts an intuitive interface. The system load within one second and prevents unauthorized access. The system capable of handling the increasing number of users. With scalability in mind, the project lays a foundation for expanding to other universities or communities in the region, reinforcing the role of technology in advancing women’s empowerment and autonomy in the automotive sector.",information systems,2024 – 2025,"Community-driven support, Artificial intelligence, Content moderation, Decision-making autonomy",67
+P062,Kidfuel,"Internationally, it is well established that the behavior, performance, and achievement of children are directly linked to the nutritional status of their overall diet. With working parents and a variety of responsibilities, the content of the children’s lunchboxes is often neglected or replaced with processed and prepackaged food. The increase in obesity and overweight children cause health problems and a decrease in academic performance indicates a solution is needed to resolve this issue. “KidFuel” is an application that aims to encourage children to eat healthy food and motivate parents to promote this lifestyle by providing flexible subscriptions, multiple diet options, rewards program. This application promotes a healthy lifestyle, giving parents enough flexibility and options while saving their time and the hassle of meal planning.",information systems,2024 – 2025,"Kidfuel, Lunchboxes, Obesity, Overweight, Diet options.",68
+P063,MAREEN – مَـرین Game-Based Physical Therapy Using AR and Body Tracking for Children’s Fracture Recovery,"Temporary disabilities, such as fractures, are common injuries that can affect humans. The legs and arms are particularly vulnerable to these injuries, especially in children, which necessitate physical therapy sessions after their casts are removed. However, children face difficulties in adhering to these sessions, due to their quick boredom with traditional exercises and the high cost of therapy sessions, which burdens parents and affects the continuity of treatment. Mareen offers a solution to these challenges through an innovative digital game that transforms physical therapy into an engaging and accessible experience. The game uses augmented reality, body tracking, and rule-based programming to deliver interactive sessions and provide immediate feedback on the child's performance. With today's digital generation accustomed to using smart devices and interactive games, integrating therapy with playful elements is an effective way to engage and motivate them to continue. The game reinforces the child's commitment through a points and rewards system, and a fun experience that includes an egg that hatches into a virtual animal upon completion of all exercises, adding an exciting and fun touch to the treatment journey.",information systems,2024 – 2025,"Physical therapy, Augmented reality, Body tracking, Rule-based programming, Fractures.",69
+P064,Be Prepared Game,"Saudi Arabia has experienced a rise of gaming interest in the past few years, fueled by its rapidly expanding number of users and giant investments in the video game ecosystem. In this vibrant environment, games can serve as powerful educational tools. Extreme weather conditions in different parts of Saudi Arabia can cause emergency situations for members of general public. The goal of this project is to develop an interactive educational game that would help to create more awareness about coping with different weather conditions, these include snowfall, high temperature and flooding around the kingdom of Saudi Arabia. Understanding local weather patterns and their associated risks is crucial for ensuring safety and preparedness. The game is targeted at children aged 7 to 16. The game allows players to navigate various weather scenarios, make quick decisions to gather essential resources, and respond to challenges. This interactive approach not only educates but also empowers young players with the knowledge needed to handle weather-related situations effectively. Incorporating local cultural references makes the game relatable, while mini challenges reinforce learning about safety in problematic weather conditions. The game utilizes modern design methods, providing an easy-to-use interface and a smooth user experience. A structured requirements analysis was conducted to ensure the final product meets the needs and expectations of its target audience. The game was implemented using Unity and C#, featuring three levels based on heatwaves, snowfall, and rain, with unique scenarios for both boy and girl characters. Testing included functional and non-functional testing to ensure the game’s functionality, stability, and ease of use. Feedback helped improve gameplay and enhance the overall user experience.",information systems,2024 – 2025,"Educational games, Game Development, Unity",70
+P065,محيط Color-Coded Emotional Learning: An AI Model for Children with Autism,"Autism spectrum disorder (ASD) affects one in every thirty-six children worldwide, making it one of the most prevalent cognitive disabilities. Classified as a neurodevelopmental disorder, ASD impairs an individual’s ability to recognize and respond to emotions in social interactions. Difficulties in understanding their own emotions or those of others can significantly impact a child's development and communication skills. To address this challenge, the Muheet application introduces a novel approach to emotion perception by leveraging synesthesia, enabling individuals to associate emotions with specific colors. This is achieved through interactive games designed to train the mind in recognizing emotions through logically assigned color associations. Additionally, AI-based facial expression recognition validates this process, ensuring accuracy and engagement. Developed using agile methodology and designed with accessibility guidelines in mind, Muheet ensures inclusivity for all users. A survey conducted to evaluate its effectiveness revealed that 78% of participants expressed confidence in Muheet’s methodology and its potential to improve emotional communication for children with autism.",information systems,2024 – 2025,"Autism, Facial Expression Recognition, Machine Learning, Accessibility",71
+P066,AllWay,"Urban mobility is essential in rapidly growing cities like Riyadh, where the need for accessible and sustainable transportation solutions is critical. This thesis focuses on the AllWay mobile application, designed to enhance walking, running, and hiking experiences in Riyadh by providing detailed route planning and user-centric features. Riyadh’s extensive urban landscape and reliance on motorized transportation present significant barriers for pedestrians and tourists seeking to explore the city on foot. Current navigation apps primarily cater to vehicular traffic, leaving non-motorized travelers underserved. The AllWay app addresses this gap by offering a platform tailored to walking, running, and hiking routes, integrating route customization, and information on nearby amenities and landmarks. Through user surveys and research, this study identifies key needs for accessible routes, personalized preferences, and social features. The app incorporates features such as difficulty level filtering, community feedback, and integration of local cultural sites, enhancing both physical activity and urban exploration. AllWay is aligned with Saudi Vision 2030’s goals of sustainable urban development and active mobility. The app’s design incorporates features to provide an intuitive user experience, promoting healthier lifestyles and sustainable tourism. The findings suggest that AllWay can significantly improve urban mobility in Riyadh, fostering active engagement among residents and visitors while supporting Riyadh's broader urban development goals. The application represents a critical step toward addressing the challenges of urban exploration and mobility, setting the stage for future innovations in route planning for non-motorized travelers.",information systems,2024 – 2025,"AllWay, urban mobility, Vision 2030, Riyadh, sustainable tourism",72
+P067,CAMPUSGUIDE (FAQ Chatbot System Using Structured Databases),"This research showcases the creation of an AI-powered chatbot system designed to improve assistance at Princess Nourah bint Abdulrahman University (PNU). By leveraging Natural Language Processing (NLP) and Artificial Intelligence (AI), the initiative aims to address challenges students face when seeking guidance outside regular office hours, offering round-the-clock support with features such as understanding user intentions and adapting responses based on input for accurate interactions. The system is built using Flutter for cross-platform accessibility, with Firebase providing backend support and integrated services such as authentication, cloud functions, and real-time data handling, and advanced AI capabilities powered by IBM Watson technology. Unlike chatbots from institutions like St. John’s University, Loyola University Chicago, and universities in the UAE and Sharjah, the PNU chatbot incorporates cutting-edge solutions to address issues like data reliance and handling complex queries. Tailored to university functions, the chatbot ensures timely updates, scalability during peak usage, and enhanced user satisfaction, improving academic service efficiency and reducing the burden on personnel. Efficient while easing the burden on personnel.",information systems,2024 – 2025,"Chatbot, FAQ, academic activities, Natural Language Processing, Artificial Intelligence.",73
+P068,MALI – Expense Tracker Application,"Managing personal finances is becoming more difficult in the digital age, with the decline of cash transactions and the rise of digital payment methods. Many individuals struggle to consistently track their expenses, leading to overspending, missed savings opportunities, and financial distress. Existing methods are often tedious and lack real-time insights, discouraging users from maintaining a clear picture of their financial habits. The Mali application offers a comprehensive expense-tracking solution. By automating expense recording through SMS integration and providing real-time insights through digital dashboards, Mali simplifies financial management to encourage informed decision-making. With its user-friendly design and practical features, Mali empowers users to achieve their financial goals and improve their overall economic well-being.",information systems,2024 – 2025,"Mali, Expense Tracker, Digital Dashboard, Digital Payments, Financial distress",74
+P069,Bridging Communication Gaps: A system for Arabic sign language translation,"Sign language plays a critical role in communication for the Deaf community, ensuring access to education, media, and social activities. The Arabic-speaking Deaf community faces significant challenges in accessing multimedia content due to the limited availability of systems that can translate Arabic subtitles into Arabic Sign Language. This gap is particularly widening considering the linguistic and cultural uniqueness of ArSL with its regional variations, among which SSL stands out. Such a challenge encourages the current study to develop a system enabling the translation of Arabic subtitles into matching ArSL gestures through advanced techniques of deep learning. The CNN, Capsule Networks, and ViT have been used along with robust preprocessing, data augmentation, and feature engineering to achieve high accuracy and adaptability. These results will help further show this system's capabilities in producing accurate and culturally relevant gestures that help bridge the gap in the accessibility of the Deaf community. This work increases accessibility to multimedia content and also makes it more inclusive and equitable, paving the way for future advancements in real-time sign language translation",Data Science and Analysis,2024 – 2025,"Arabic Sign Language (ArSL), Deep Learning, Vision Transformers (ViT), Capsule Networks, Convolutional Neural Networks (CNNs).",76
+P070,Mama’s Hope,"Preeclampsia is a life-threatening pregnancy complication that remains difficult to detect early due to its complex and multifactorial etiology. Timely identification is essential to reduce risks to both maternal and fetal health. This study presents PreDx (Preeclampsia Diagnosis), a machine learning-based predictive model developed to support early detection of preeclampsia by leveraging clinical and biological features. Using a combined dataset of 42,965 maternal health records from Yonsei University Hospital (Korea) and 19,905 records from King Abdullah University Hospital (KAAUH). This research developed a stacking ensemble model that integrates three base learners: Neural Network, XGBoost, and Logistic Regression. The proposed model achieved an accuracy of 96% and an F1-score of 0.96, reflecting a balanced performance in terms of both precision and recall. The strength of PreDx lies in its ability to reduce misclassification, especially false negatives—by systematically analyzing key risk factors. These results highlight the model’s potential to assist clinicians in early diagnosis, contributing to improved maternal care and safer pregnancy outcomes.",Data Science and Analysis,2024 – 2025,"Preeclampsia, Machine learning, biological features, Maternal health, Stacking ensemble model.",77
+P071,AI-Powered Platform for Monitoring and Analyzing Consumer Sentiments in the Cosmetics Industry,"This project aims to develop a user-friendly web platform tailored for the skincare and cosmetics industry, enabling comprehensive, data-driven insights. The platform includes an integrated search engine that allows users to find products using keywords or hashtags, connecting directly to a text-based social media platform, such as Twitter. The system is designed to gather and analyze tweets related to the searched products, performing sentiment analysis to provide users with an overall understanding of public sentiment. This analysis helps highlight customer opinions and trends that are critical for informed decision-making. To enrich user experience, the platform employs generative AI to generate concise textual summaries, showcasing the main sentiments and themes mentioned in social media discussions. This feature allows users to quickly capture the essence of public perception without needing to sift through numerous posts. The project also incorporates interactive visualization tools, presenting key sentiment data and analytics on a dashboard. This visual interface makes it easy for users to interpret data, supporting cosmetic marketing teams in tracking consumer feedback and conducting competitor analysis. By providing these insights in an accessible format, the platform empowers businesses to refine their marketing strategies and enhance product evaluation",Data Science and Analysis,2024 – 2025,"Skincsre, Cosmetics, Twitter scraper, Product review, Sentiment analysis",78
+P072,AutoReview,"The Saudi Food and Drug Authority (SFDA) currently relies on domain experts to manually review Summaries of Product Characteristics (SPCs) submitted by pharmaceutical companies. Manual comparisons between generic drugs and registered New Chemical Entities (NCEs) are time-consuming and prone to human error, potentially impacting the efficiency and reliability of regulatory decisions. This project present AutoReview, an AI-based system proposed to completely automate the SPC review process to alleviate these issues. The proposed system leverages transformer-based models in the automatic documentation of contextual comparison and data extraction. Our approach uses cosine similarity to calculate the semantic similarity between sections of the SPC. To improve transparency and interpretability, AutoReview utilizes the thresholds based highlighting approach aimed at greater transparency and interpretability, producing color-coded results that clearly mark sections as either compatible or incompatible. The framework acts in alignment with principles of explainable information retrieval. Reviewed by experts, AutoReview aims to outperform existing solutions in maintaining internal consistency, increasing automated workflows, and rationalizing regulatory analysis.",Data Science and Analysis,2024 – 2025,"Saudi Food and Drug Authority (SFDA), Summaries of Product Characteristics (SPCs), Natural Language Processing (NLP), Transformer, Interpretability.",79
+P073,Analysis of Ancient Rock Inscriptions in Saudi Arabia Using Machine Learning,"Saudi Arabia is home to numerous archaeological sites featuring ancient rock inscriptions, many dating back thousands of years. These inscriptions offer invaluable insights into the region’s historical, cultural, and linguistic heritage. They serve as a crucial link in understanding the evolution of the Arabic language and its integration into the broader context of the Arabian Peninsula. Despite their significance, this rich archaeological record remains relatively underexplored, particularly in relation to the application of artificial intelligence and machine learning. Aligned with Saudi Arabia’s Vision 2030 objectives to preserve and promote cultural heritage, this project presents Thamad. This innovative application employs AI techniques to detect and translate Thamudic inscriptions into Arabic. Developed in collaboration with the Heritage Commission, the system utilizes annotated image data processed through object detection and Optical Character Recognition (OCR) models to locate inscriptions and recognize individual characters photographed by users. In addition to its translation capabilities, Thamad provides an interactive platform for language learning through educational games, encouraging broader public engagement with the ancient script. By integrating advanced technological tools with domain-specific knowledge, this project contributes to cultural preservation while enabling a deeper understanding of Saudi Arabia’s archaeological legacy.",Data Science and Analysis,2024 – 2025,"Ancient rocks’ inscription, Saudi Arabia heritage, AI, image recognition, Image translation.",80
+P074,بصيرة Baseera: An AI-Based Sound Recognition and Visual Alert System for the Deaf,"Individuals with hearing impairments often encounter significant barriers in recognizing environmental sounds that are essential for personal safety and situational awareness. To address this challenge, this project introduces Baseera — an AI-driven sound recognition and alert system designed as a mobile application to support deaf and hard-of-hearing users. The application focuses on detecting and classifying five critical emergency-related sounds: fire alarms, sirens, gunshots, dog barks, and car horns. Upon detection, Baseera immediately notifies users through intuitive visual and tactile alerts to ensure timely awareness. The development of Baseera followed a structured machine learning pipeline, beginning with comprehensive data preprocessing and the extraction of Mel-frequency Cepstral Coefficients (MFCCs) as sound features. Multiple classification algorithms were explored, including Support Vector Machines (SVM), K-Nearest Neighbors (KNN), Random Forest (RF), and a one-dimensional Convolutional Neural Network (1D-CNN). Experimental evaluation demonstrated that the 1D-CNN model achieved superior performance, attaining a test accuracy of 94% in accurately classifying the targeted sound categories. The final system was integrated into a mobile application using Flutter for the Android front-end and Flask for the back-end, enabling real-time sound classification and instant notification delivery. By translating critical auditory information into accessible visual and haptic feedback, Baseera has the potential to enhance the safety, independence, and situational awareness of individuals with hearing impairments, promoting a more inclusive and responsive digital environment.",Data Science and Analysis,2024 – 2025,"Sound Recognition, Deaf Accessibility, Machine Learning (ML), Deep Learning (DL), MFCC",81
+P075,RASAD - An AI-Powered Safety Detection System for Monitoring and Ensuring Compliance in Restaurants and Kitchens,"The food service industry continues to face significant challenges in maintaining hygiene standards, primarily due to reliance on outdated and labor-intensive manual monitoring practices, which increase the risk of health violations. In Saudi Arabia, recent reports indicate that over 50% of foodborne illness cases are directly attributable to safety non-compliance within restaurants [Source needed]. This project introduces Rasad, an intelligent, automated system designed to monitor and enforce food safety compliance through computer vision and machine learning techniques. The development of Rasad followed a structured machine learning pipeline, beginning with rigorous image preprocessing to enhance model robustness and accuracy. Several state-of-the-art object detection algorithms were systematically evaluated, including Region-based Convolutional Neural Network (R-CNN), You only Look version 5 (YOLOv5), YOLOv8, and the emerging YOLOv11. Experimental results demonstrated that YOLOv11 outperformed its counterparts, achieving a mean Average Precision at 50% IoU (mAP50) of 86.7% in detecting common hygiene violations, such as the absence of gloves, masks, or hairnets in restaurant kitchen environments. The optimized YOLOv11 model was integrated into a user-friendly web application built with Streamlit, enabling real-time analysis of image and video feeds from surveillance cameras. By automating the detection of safety breaches, the Rasad system offers a proactive solution to mitigate hygiene violations at their source. Beyond its technical contributions, Rasad holds the potential to significantly enhance public health outcomes, reduce the regulatory burden on health authorities, and promote a culture of accountability and safety in Saudi Arabia’s food service sector.",Data Science and Analysis,2024 – 2025,"Object Detection, You Only Look Once (YOLO), Safety Compliance, Real-Time Monitoring, AI Integration",82
+P076,AOUNBOT: RAG-Based Virtual Assistant for Tourism in Saudi,"Aligned with Saudi Arabia’s Vision 2030, the tourism and entertainment sectors are driving economic diversification. Despite the rapid growth of events, festivals, and cultural performances, residents and visitors still face challenges accessing centralized, reliable information about events, festivals, and cultural attractions. This project introduces AOUNBOT, a Gen AI-powered virtual assistant that leverages Retrieval-Augmented Generation (RAG) to provide accurate, real-time updates on entertainment events and museums across the Kingdom through a user-friendly interface. A dataset compiled from various sources served as the knowledge base for the research. The study evaluated the all-MiniLM-L6-v2 embedding model and the generation models GPT-4-turbo, GPT-4o, FLAN-T5 Base, and FLAN-T5 Large. The fine-tuned FLAN-T5 Base delivered the best performance, achieving an F1 score of 0.81, precision of 0.84, recall of 0.81 and MMR of 0.58. The overall effectiveness of the chatbot was assessed through a user survey, evaluating criteria such as helpfulness, accuracy, and coherence, with the results indicated generally good performance across all aspects. This research emphasizes the use of RAG technology to optimize Gen AI models in the tourism domain, enhancing visitor satisfaction and positioning Saudi Arabia as a leading global tourism destination.",Data Science and Analysis,2024 – 2025,"Retrieval-Augmented Generation (RAG), Generative AI, Virtual Assistant, Tourism, Entertainment",83
+P077,Medicinal Product Classification Model For SFDA,"The objective of this project is to automate the current lengthy and time-consuming medicinal product approval process of the Saudi Food and Drug Authority (SFDA) by applying Deep learning techniques to automatically classify products as either prescription or over-the-counter (OTC). The CRISP-DM framework is used to analyze data from SFDA’s Human Drugs file and Summary of Product Characteristics (SPC) files. The Medicinal Product Classification Model was developed using Natural Language Processing (NLP) and Deep learning (DL) techniques. Various models were explored to ensure efficient and accurate classification of medicinal products based on Summary of Product Characteristics (SPC) files. The proposed models were trained, validated, and tested, with evaluation metrics such as accuracy, precision, F1-score, recall, AUC-ROC, and confusion matrix deployed to assess performance. The expected outcome of this project is to build a web-based machine learning system capable of automating the classification process, enhancing efficiency and regulatory compliance for the Saudi Food and Drug Authority (SFDA).",Data Science and Analysis,2024 – 2025,"medicinal product classification, Deep learning, natural language processing, OTC, prescription.",84
+P078,وَدَق AI-Driven Serious Games for ADHD Treatment,"Attention-deficit hyperactivity disorder (ADHD) is a neurodevelop mental condition that significantly affects daily functioning, academic and workplace productivity. The unique challenges that individuals with ADHD face in professional settings are often irresolvable through traditional treatments, thereby limiting these populations’ ability to focus, manage tasks and maintain efficiency at work. To address this problem, we developed an AI-driven gamification tool called WADAQ, which is designed to enhance cognitive and executive functioning. Using machine learning (ML) models, WADAQ provides a personalised and engaging approach to ADHDsupport. Initially, classification models assess and predict ADHDlikelihood with high accuracy, after which users undergo subtype assessment to enable the system to recommend tailored mini-games on the basis of reinforcement learning (RL) techniques. These features are intended to help individuals develop key daily life skills, such as sustained attention, impulse control and task management. An interactive dashboard enables real-time performance tracking for optimized training. The initial results indicated that when WADAQ leverages a support vector machine (SVM), it achieves 97% predictive accuracy. This accuracy is maintained after hyperparameter optimisation with GridSearchCV.",Data Science and Analysis,2024 – 2025,"ADHD, Game, Machine Learning, Reinforcement Learning, Accessibility",85
+P079,A Software-Defined Networking Approach to Link Failure Solutions in Cloud Environments,"Cloud computing is an important part in the information technology field, with users heavily dependent on its services, ensuring reliable connectivity is essential, especially in addressing link failures that can disrupt services in the network. Therefore, the need for efficient techniques that handle failure recovery in networks is increasing. This project introduces a link failure recovery approach via Software Defined Networking (SDN) that focuses on detecting and managing link failure in the network using different mechanisms, which leverage the centralized control and overall view and management of the network due to the nature of SDN, this leads to simplifying networking complexities, faster rerouting, automated failure detection, and improving the network performance. We implement and evaluate our proposed failure recovery mechanism using the ONOS SDN controller and Mininet network emulator. We test a Google WAN-style topology consisting of 12 switches and 24 hosts to simulate realistic wide-area network scenarios. We connect the Mininet topology to ONOS and activate reactive, proactive, and hybrid forwarding mechanisms. We test two failure scenarios (3 and 6 link failures) and evaluate performance metrics including bandwidth and latency (RTT). Our results show that the hybrid forwarding mechanism outperforms both proactive and reactive approaches by maintaining higher bandwidth and lower latency during failures. The system meets all defined functional and non-functional requirements, demonstrating high reliability, scalability, and performance in handling link failures in SDN environments.",Information Technology,2024 – 2025,"Cloud computing, Link failure, Software defined networking, ONOS, Mininet.",87
+P080,A Smart Mobile Application For Skin Cancer Detection,"Developing a mobile application for skin cancer is crucial for the early detection of the disease. On the other hand, the mobile application will educate users about skin changes. Therefore, this project focuses on the design of a mobile application for the early detection of skin cancer in the Kingdom of Saudi Arabia (KSA), where skin cancer has become a public health issue owing to high levels of ultraviolet radiation. The main goal is to facilitate the early detection of skin cancer by simply enabling users to upload photos of their skin lesions. Moreover, another module based on Artificial Intelligence (AI) will be responsible for analyzing the uploaded image. It relies on a Convolution Neural Network trained on the PAD-UFES-20 dataset to predict if the uploaded image contains any type of skin cancer. Thus, the main features of your system consist of providing accurate and quick analysis of skin lesions.",Information Technology,2024 – 2025,"Skin cancer detection, Convolutional Neural Networks (CNN), Deep Learning (DL), Mobile application, Artificial Intelligence (AI).",88
+P081,تنبا Artificial Intelligence for Predicting Stock Market Prices,"This project proposes a stock price prediction system for three Saudi companies including Zain, Dar Al Arkan, and Salamah, based on machine learning (ML) and deep learning (DL) techniques. After applying different algorithms for both long and short-term forecasting, the Long Short-Term Memory (LSTM) model demonstrated the best performance for short-term prediction with an R2 value of 0.91 and RMSE of 0.02, while the support vector regression (SVR) was most effective for long-term trends, achieving an R2 of 0.99 and an RMSE of 0.07. To improve accuracy, the models incorporate technical indicators such as the Moving Averages (MA) and the Relative Strength Index (RSI) and are trained on 10 years of historical data from Tadawul. The system is designed as a web-based platform that helps investors make informed decisions based on AI-driven insights. By leveraging advanced AI techniques, this project contributes to the growth of financial technology in Saudi Arabia.",Information Technology,2024 – 2025,"Stock Market, Machine Learning, Deep Learning.",89
+P082,Predicting Diabetes Patients Based On Machine Learning Techniques,"Our graduation project applied machine learning algorithms to develop a reliable model for diabetes prediction. Using a dataset of 1,000 patient records with attributes such as age, gender, BMI, cholesterol, and creatinine levels, the project identified patterns that suggested potential diabetes risk. By integrating patient data with medical test results, the model aimed to assist healthcare professionals in early diagnosis, supporting informed treatment planning.Machine learning brings transformative potential to the tech and healthcare fields by automating data analysis and enhancing predictive accuracy, particularly for complex conditions like diabetes. In this project, we incorporated various machine learning algorithms, including Decision Trees, Random Forest, Logistic Regression, Support Vector Machines (SVM), and Neural Networks. Each algorithm was chosen for its specific advantages: Decision Trees and Random Forests offer interpretable, rule-based outputs that make predictions transparent and actionable for clinicians. SVM and Neural Networks excel at identifying complex, non-linear patterns in large datasets, making the model highly adaptable and effective in recognizing subtle risk factors. Logistic Regression contributes by efficiently managing binary classification tasks, which align well with the predictive needs of healthcare applications.The model's reliability was enhanced by comparing the algorithms' performance across various metrics such as accuracy, precision, recall, and F1-score and feature selection methods were selected to refine the model by identifying key predictive features, like BMI and HbA1c levels. This approach demonstrated how machine learning can transform raw medical data into actionable insights, delivering reliable and efficient predictions that support healthcare providers in identifying high-risk patients early and providing the patients with the needed healthcare plan early.",Information Technology,2024 – 2025,"Machine learning, diabetes prediction, comparative analysis",90
+P083,Remote Delivery of Healthcare Services for Managing Heart Failure,"Heart failure continues to pose a significant worldwide health problem owing to its high death rates and high rates of readmissions to hospitals. In reaction, this project proposes a functional web-based system called HeartEverBeat aimed at remotely tracking heart failure patients and aiding in their continuous treatment. The system allows the registration of the patients, the handling of their profiles, and the direct transmission of essential health data including symptoms, vital signs, and medical history to physicians. Doctors can access this data, prescribe medications and treatment plans, and maintain communication with patients via integrated chat and call features. The system also includes roles for Admin and Super Admin, allowing for efficient user management and announcement broadcasts. The system was developed using Agile development and extensively tested in every module to guarantee ease of use, role-based security, and dependability. The system facilitates real-time communication, continuity of care, and minimizes the need for in-person visits especially valuable during crises like the COVID-19 pandemic. Through this development, we aim to enhance remote healthcare delivery, promote improved patient outcomes, and set the stage for future integration of wearable technology, artificial-intelligence-informed analytics, and support for multiple languages.",Information Technology,2024 – 2025,"Telehealth, Heart Failure, Diagnosis, symptoms, treatment plan.",91
+P084,AI-Powered System for PGS Data Management and Analytics,"The existing manual procedures result in inefficiencies, inaccuracies in data entry, and a lack of real-time information access. The new system will consolidate data storage, simplify administrative tasks, and utilize artificial intelligence for enhanced data analysis and reporting. Featuring an intuitive web interface, it will ensure seamless usability for administrators, faculty, and students. By automating processes and providing intelligent insights, the system will enable data-driven decisions while improving the reliability and availability of postgraduate data. The system's architecture consists of three core components: a robust database application for secure data storage and management, an AI-powered module for advanced analytics and automation, and an intuitive user interface designed for seamless interaction. This solution is designed to boost operational efficiency, increase transparency, and elevate the overall quality of postgraduate program management. Additionally, it aligns with the university’s digital transformation objectives, fostering continuous improvement and quality assurance in postgraduate education.",Information Technology,2024 – 2025,"Postgraduate studies, Data Management and Analytics, Artificial Intelligence",92
+P085,Driver drowsiness detection system,"Drowsy driving is one of the major causes of road accidents, especially on long highways and during night travel. Many drivers are unaware that they are sleepy until it is too late. This project aims to solve that problem by developing a real-time drowsiness detection system using a smart camera (Nicla Vision) and a lightweight AI model. The system monitors the driver’s face and checks if their eyes are closed. If the eyes stay closed for more than three seconds, a buzzer sounds to alert the driver. We trained a custom machine learning model using images of eyes labeled as “Drowsy” and “Not Drowsy” through Edge Impulse. The model was then deployed to the Nicla Vision board, allowing it to run completely offline and in real time. The use of a small embedded device ensures low cost, energy efficiency, and easy installation in any vehicle. The prototype was tested under controlled conditions and successfully detected signs of drowsiness with 95% accuracy. This solution is practical, affordable. It highlights how AI and embedded systems can improve public safety. By alerting drowsy drivers early, the system helps reduce accident risks and contributes to building safer roads.",Information Technology,2024 – 2025,"Drowsy Driving, Driver Monitoring, AI, Nicla Vision, Real-Time Alert System",93
+P086,عَون AWN Volunteer management website,"Volunteering is the act of individuals offering their time, skills, and resources to benefit others without financial compensation, enhancing the cohesion and progress of society. To support this field, we are developing an intermediary website that connects volunteer organizations with volunteers, making it easier for them to collaborate effectively. Our website simplifies volunteer program management by helping volunteers or guests track attendance, record and calculate volunteer hours, and view detailed program plans. Supervisors can assign specific tasks and responsibilities, while progress reports are automatically sent to all parties involved to ensure transparency and accountability. Additionally, certificates are automatically issued to volunteers, personalized with their names and achievements, and electronically signed by the supervisor. This solution will streamline the volunteer experience, making it more efficient and rewarding for both volunteers and organizations.",Information Technology,2024 – 2025,"Volunteering, Volunteer organizations, Volunteer management, Attendance tracking, Volunteer hours",94
+P087,TAWASUL: AI Application for Arabic Language Translation,"Due to the rapid development of artificial intelligence technologies in recent years, new frontiers have opened regarding better communication across diverse languages. However, issues persist, especially for Arabic, with its unique set of linguistic characteristics. This leads to a number of obstacles that further impede effective communication and maintain barriers in sectors such as customer service, tourism, and politics, where effective communication is essential. This project introduces ""Tawasul,"" an innovative application capable of detecting Saudi dialect and translating with accuracy into English, hence allowing cross-cultural communications easily. Using the latest NLP and AI techniques to facilitate accurate detection and translation, with an initial focus on the Saudi dialect. In global communication, Tawasul develops as a bridge between worlds, transcending linguistic frontiers and making available at every turn what has become the very complicated landscape of Arabic dialects. This new application is more than a translation tool-it connects Saudi and English-speakers in meaningful and accurate ways through the power of artificial intelligence and NLP. Tawasul goes even deeper, into the subtlety of the language and the richness of context in Arabic communication. This application represents a deep commitment to the breaking down of barriers in communication. Tawasul is not just a technological solution, it is a vision of mutual understanding, but rather as a window into each other's worlds, fostering deeper, more authentic cross-cultural dialogues that celebrate our shared humanity.",Information Technology,2024 – 2025,"Artificial Intelligence, Natural Language Processing, Speech Recognition, Arabic language, Translation",95
+P088,20-EIGHT: Website for Arabic language learning,"Technology has inspired us and facilitated many difficulties in learning new skills. As a result, learning new language websites must adapt technological advancements to provide compatible websites, especially in Arabic, due to a lack of sufficient resources. We have noticed that finding resources for learning the Arabic language easily and at any time requires the development of smart websites. This project aims to help learners worldwide get a comprehensive and attractive educational experience in the Arabic language. The website will contain different user roles, including learner, teacher, and administrator with a user-friendly interface and robust functions. Sufficient types of courses are available to support a wide range of learners, from beginners to advanced. The website will allow learners to view and book available slots based on teacher schedules. By providing a connection between learners and efficient teachers through scheduled live lessons the website will be a powerful tool for learners looking to improve their speaking skills. Also, the integration of AI assists learners in doing their homework and provides instant corrections. This project will be implemented using Django back-end, and React.js front-end. As a result, learners will be able to improve their Arabic skills by accessing the website at any time to learn and practice, ensuring continuous learning.",Information Technology,2024 – 2025,"E-Learning, Arabic language, Interactive Classes, AI-Assistant, Live Sessions",96
+P089,AI-Powered Platform for Plant Identification and Diagnosis,"LeafNet is an innovative web application utilizing AI to provide comprehensive plant health solutions. Leveraging deep learning, it diagnoses plant diseases from uploaded images and features an AI chatbot expert for plant care inquiries. This project addresses accurate disease identification and expert guidance, integrating image processing, machine learning, and conversational AI. This report details LeafNet's design, functionality, architecture, implementation, testing plan, evaluation, and final results. LeafNet offers a dynamic resource for plant lovers, empowering informed decisions and contributing to sustainable plant care.",Information Technology,2024 – 2025,"plant health monitoring, deep learning, artificial intelligence, agriculture technology, leaf disease,",97
+P090,DeepSight: Video Deepfake Detection System,"The rise of computation power has recently resulted in new techniques and superior tools for generating hard-to-determine and convenient videos “recognized recently as Deepfakes” for several purposes. Advertisement, entertainment, and educational uses are considered as ethical deepfakes. Unfortunately, malicious users can misuse these tools for unethical reasons, such as spreading political distress, creating fake evidence, and extorting innocent people. The widespread disinformation necessitates the responsibility to prevent the spread of deepfakes and reduce their threats. The proposed DeepSight system is designed to accurately detect deepfakes by leveraging combination of machine learning and deep learning techniques. DeepSight processes videos by focusing on the face area and classifying whether a video is authentic or manipulated. The system employs a deep-learning-based feature engineering approach that integrates the advanced capabilities of YOLO-Face and Xception in feature extraction. Global Average Pooling (GAP) will select the optimal feature set. Those informative feature sets are used to train Support Vector Machine (SVM) and Long Short-Term Memory (LSTM) classifiers. Deepfake Detection Challenge (DFDC) and FaceForensics++ (FF++) datasets are used to achieve reliable and effective detection results. After testing and evaluation, the SVM model achieved an Area Under the Curve (AUC) of 96.2%, while the LSTM reached an AUC of 93.2%.",Information Technology,2024 – 2025,"Deepfake, Face Manipulation, Artificial Intelligence, Machine Learning, Deep Learning.",98
+P091,Mail Pawns: Advanced Email Malicious Detection System,"This project addresses the increasing risk of malicious emails by creating a bilingual detection system designed to improve the safety via email. Mail Pawns aims to enhance detection especially in Arabic, which is an area that is frequently overlooked in current solutions. By allowing users to manually identify suspicious emails and take appropriate action, and equipping admins with monitoring tools, the system ensures a balance between usability and control. The solution was developed with a focus on critical needs, including the increasing frequency of email attacks, the importance of supporting Arabic language detection, and the desire to streamline user experience. Ultimately, this project advances the safety and reliability of communication platforms by applying intelligent design, machine learning techniques, and language-specific natural language",Information Technology,2024 – 2025,"Malicious emails,Bert, Natural language processing,Machine learning,Artificial intelligence",99
+P092,Deep Learning for Credit Card Fraud Detection improvement (FraudShield),"Payment cards offer a convenient method for making purchases. The rise of fraud cases creates financial risk in the commercial sector. In this project, we aim to create a DL-based detection model for card fraud detection using both publicly available and real transaction records. Appropriate performance metrics will be evaluated on real-world datasets. Advanced deep learning techniques will be utilized in the project to identify fraudulent patterns effectively. Much emphasis will be given to handling imbalanced datasets, a common challenge in fraud detection. Strong data preprocessing and feature engineering strategies will be adopted to enhance model accuracy. This approach aims at contributing toward a scalable and reliable fraud detection solution.",Information Technology,2024 – 2025,"Credit Card Fraud, Deep Learning, Anomaly Detection, Imbalanced Data Handling, Financial Security.",100
+P093,Deep Learning Based Wound Classification And Treatments Suggestion,"This project aims to develop an automated system for classifying wound types using deep learning techniques. The system relies on Convolutional Neural Networks (CNNs) to analyze wound images uploaded by patients or healthcare providers, identifying the wound type accurately and suggesting appropriate treatment recommendations. The motivation behind this work is to reduce reliance on manual wound assessment, which can vary depending on the experience of healthcare professionals, and to provide a fast, reliable, and accurate diagnostic tool. The system features a simple user interface that allows users to upload images. These images undergo preprocessing before being analyzed by the model, which outputs the predicted wound type along with corresponding treatment suggestions. The project addresses challenges such as variations in image quality and limited training data, aiming to improve classification accuracy and support better clinical decision-making. Additionally, the system has potential applications in telemedicine and digital healthcareservices, contributing to the advancement of remote medical care solutions.",Information Technology,2024 – 2025,"Wound Classification, Deep Learning, Convolutional Neural Networks (CNNs) and Artificial Intelligence.",101
+P094,Enhanced Brain tumor detection and classification based on Deep learning,"Detecting and diagnosing brain tumors is challenging and critically important, they frequently take a long time and heavily rely on a skilled interpretation of the images using MRI scans. This project goal is to support healthcare professionals by developing a system that can automatically identify and categorize brain tumors using deep learning. By fine-tuning an advanced, pre-trained model, we concentrate on helping the system to distinguish between various types of tumors. To ensure its accuracy and flexibility, we use techniques that expand the dataset by adding other images, hence enhancing the model's ability to handle particular situations. The system not only recognizes the existence of a tumor but also classifies it, giving doctors precise predictions and verifiable results to review. Finally, by providing radiologists with immediate and measurable results, this project may help them make more accurate and timely treatment decisions for their patients.",Information Technology,2024 – 2025,"Brain tumor detection, Deep learning, MRI scans, Medical imaging, Tumor classification",102
+P095,Image Forgery Detection Website: ForgeFinder,"As digital manipulation software became available and the spread of misinformation became widespread, authenticating the genuineness of digital images was an essential need across industries. Forged images in social media, journalism, and legal investigations falsify facts, frame public opinions, and discredit evidence. Preempting this growing issue, ForgeFinder was developed as a smart, real-time web application that identifies image forgery with high precision and speed. ForgeFinder uses a combination of forensic analysis tools, including Error Level Analysis (ELA), to look for inconsistencies in image compression levels—typically a sign of image manipulation. ForgeFinder preprocesses and analyzes uploaded images using optimized image processing pipelines, delivering real-time results to the user in a clean and responsive web interface. The user can view detailed visual feedback and save analysis history securely in a personalized dashboard. The platform is built using Python, Flask, HTML/CSS/JS, and MySQL, and the development and test environment is managed using tools like Anaconda, MAMP, and Mailtrap. The use of platforms like Kaggle enabled the team to try out different detection models and datasets efficiently without relying on local resources. With reliability and scalability in its concept, ForgeFinder is not only a technical endeavor but a pragmatic product with pragmatic uses. It is a beneficial instrument for people, professionals, and institutions to easily and precisely establish image authenticity. Future enhancements include upgrading the system into video analysis, deeper detection accuracy, and cloud deployment for broader accessibility.",Information Technology,2024 – 2025,"Forgery Detection, ELA, CNN",103
+P096,سند SANAD : Proactive Defense for Individuals with Bipolar Disorder Through a Personlized Cyber Protection,"Bipolar disorder (BD) affects approximately 40 million individuals globally, including over 320,000 in Saudi Arabia. This condition is characterized by manic and depressive episodes, which drive impulsive behaviors such as poor decision-making and risk-taking. These behaviors expose BD patients to significant cybersecurity threats, including phishing, privacy breaches, and unauthorized financial activities. This research explores tailored cybersecurity interventions to effectively mitigate online risks for individuals with BD. Despite advancements in cybersecurity, the unique vulnerabilities of BD patients remain underexplored, with limited integration of behavioral analytics in current solutions. Existing systems fail to account for the dynamic and unpredictable behaviors associated with mood episodes, creating a gap in behavioral cybersecurity practices. Addressing this gap is essential for protecting this vulnerable population and advancing the field of cybersecurity. This study proposes Sanad, an advanced cybersecurity solution tailored to BD patients. The system leverages keystroke dynamics to monitor behavioral patterns, detect mood shifts, and implement adaptive protective measures. Core features include application and web access restrictions, financial transaction controls, adaptive trusted-contact management, and real-time monitoring mechanisms to safeguard sensitive information, ensuring that users’ privacy remains uncompromised during high-risk episodes. These measures are designed to mitigate impulsive online actions, safeguard sensitive information, and enhance user autonomy. By bridging the gap between mental health and cybersecurity, this research aligns with the Saudi National Cybersecurity Authority’s (NCA) strategic objectives to enhance behavioral cybersecurity. This research aims to establish Sanad as a pioneering solution that protects high-risk populations and contributes to the development of innovative cybersecurity practices.",Cybersecurity,2024 – 2025,"Personalized Cybersecurity, Behavioral Cybersecurity, Cyberpsychology, Bipolar disorder, Adaptive Cyber Defense",105
+P097,Hybrid Vulnerability Scanner for Detecting Known and Unknown Vulnerabilities in Web Applications (ZeroScan),"Web applications are particularly vulnerable to different types of vulnerabilities arising from both known and unknown vulnerabilities, which result in loss of data, money, and reputation. Most current scanners employ the approach based only on known patterns, or signatures, and frequently ignore unrecognized and more sophisticated forms of vulnerabilities such as zero-day vulnerabilities and context-specific anomalies. These gaps reduce their ability to detect emerging or less obvious risks that fall outside of static rule sets. To address these limitations, the proposed Hybrid Vulnerability Scanner (ZeroScan) integrates both signature-based and anomaly-based detection techniques. ZeroScan improves detection sensitivity, supports the identification of unknown vulnerabilities, and offers detailed and well-organized reporting. It is designed to be navigable by any level of user, from novice to advanced. Through the combination of hybrid detection and post-detection analysis, ZeroScan provides valuable insights to increase the level of cybersecurity of web applications in critical domains such as e-commerce, healthcare, and finance.",Cybersecurity,2024 – 2025,"Anomaly-Based Detection, Signature-Based Detection, Vulnerability Scanner, Web Application Security, Zero-Day Vulnerabilities",106
+P098,WATHEEQ Smart Contracts for Electronic Health Record (EHR) Access Control,"This project focuses on developing a secure and scalable system for managing patient health records across healthcare institutions in Saudi Arabia. It addresses the significant challenges of data privacy, security, and efficient access within a distributed environment. The project integrates blockchain technology to create a trusted decentralized solution for patient records sharing and access across Saudi hospitals. The solution utilizes Hyperledger Fabric to establish a permissioned blockchain network where each hospital maintains its own patient database, keeping local control over data while securely sharing patient data and access requests across the network. The solution uses smart contracts on the blockchain to automate the access control by verifying requester roles, ensuring only authorized entities can access patient records. The aim of the project is to enhance data security and privacy of the patient's health records through using blockchain immutable ledger and role-based smart contract mechanisms, reducing unauthorized access risks. This project aligns with Saudi Arabia’s goals to advance healthcare infrastructure by providing a secure patient health data sharing model among Saudi hospitals.",Cybersecurity,2024 – 2025,"Smart Contracts, Blockchain, EHR, Access Control",107
+P099,سد Real-Time Vishing Detection and Call Analysis System Using NLP and AI,"In an era where cyber threats are becoming increasingly sophisticated, ""سَــد"" emerges as an innovative solution to combat voice phishing (vishing) attacks, with a special focus on Arabic-speaking users. This intelligent detection system integrates Speech-to-Text (STT), Natural Language Processing (NLP), and a fine-tuned BERT (Bidirectional Encoder Representations from Transformers) model to analyze incoming calls in real time. By accurately converting spoken Arabic into written text, سَـــد detects suspicious or deceptive language patterns that indicate potential social engineering attempts. The system immediately alerts users to the risk, allowing them to make informed decisions before falling victim to scams. What sets سَـــد apart is its deep linguistic understanding of Arabic, a language known for its richness and complexity, making it highly effective in regional contexts. سَـــد not only provides reliable, real-time protection, but also raises awareness and builds user confidence in dealing with fraudulent calls. This project represents a major step forward in Arabic-language cybersecurity.",Cybersecurity,2024 – 2025,"Vishing, Artificial Intelligence, Natural Language Processing, Real time, Speech to Text",108
+P100,TOWARDS A WORLD WITHOUT PHYSICAL IDs: Face Anti-Spoof Recognition System for The Nourah Students Application.,"This project presents a robust and real-time face anti-spoofing authentication system developed for the Nourah Students application, addressing the growing security concerns in biometric verification. The proposed solution enhances the reliability of facial analysis through deep learning-based methods tailored for student identity verification. The system integrates two key components: a facial recognition module utilizing FaceNet and a spoof detection module based on the EfficientNet-B0 architecture. To improve the model's resilience against both physical and digital spoofing attacks, two data augmentation strategies-Simulated Physical Spoofing Clues (SPSC) and Simulated Digital Spoofing Clues (SDSC)-were applied during training on the LCC-FASD dataset. The model was evaluated using three distinct protocols (pl, p2.1, p2.2), simulating a variety of spoofing scenarios and demonstrating high performance, with accuracy reaching up to 94.3% and strong generalization capabilities. Despite being developed and tested under limited computational resources, the system achieved low false acceptance and rejection rates, confirming its suitability for real-time deployment. Compared to contemporary methods, this approach offers an effective balance of accuracy, efficiency, and security, making it a viable solution for integration into various applications that require trustworthy facial authentication.",Cybersecurity,2024 – 2025,"Face Anti-Spoofing, Deep Learning, Face Recognition, Real-Time Authentication, Face ID.",109
+P101,A Near Real-Time Phishing Website Detection Tool Using a Hybrid Model of Deep Learning Algorithms,"Phishing attacks represent a major cybersecurity threat, as they deceive users into disclosing sensitive information through websites that mimic legitimate ones. The increasing complexity of phishing techniques, along with limited user awareness, makes timely detection a persistent challenge, despite significant advances in security technologies. To address this issue, we present “لصَن”, a smart and efficient browser-based tool for near real-time phishing website detection. The system is powered by a custom-built hybrid deep learning modeldeveloped entirely by our team—that combines Convolutional Neural Networks (CNN) and Long ShortTerm Memory (LSTM) architectures. Our approach analyzes both URLs and webpage content, enabling more accurate and reliable detection. We designed and trained the hybrid model using a self-constructed dataset of 23,068 samples, evenly divided between 11,534 phishing and 11,534 legitimate websites. Each sample includes 24 distinct features, carefully extracted from both the URL structure and the HTML content of the page. Our hybrid model achieved an impressive accuracy of 97.79%, highlighting its effectiveness in detecting phishing website threats. By integrating the hybrid model into a browser extension, “لصَن” provides users with near real-time protection and promotes safer browsing behavior.",Cybersecurity,2024 – 2025,"Hybrid model, Near real-time website detection, URL analysis, Website content analysis, Browser extension.",110
+P102,DeepShield – Deepfake Detection Tool,"In recent years, deepfake technology has raised significant concerns regarding the authenticity and integrity of digital media. Deepfakes, which leverage development in artificial intelligence to manipulate images and videos, pose threats to privacy, cybersecurity, and social trust. This project focuses on developing a robust deepfake detection tool using deep learning algorithms. Our approach introduces an input validation layer that pre-processes the user-uploaded media (image or video) to ensure it meets quality standards and is free from inconsistencies or distortions that could affect the detection results, thereby enhancing accuracy. The verified media is then analyzed by a deep-learning detection model that will be trained on diverse datasets of real and manipulated media. This project highlights the challenges and breakthroughs in combating deepfakes while proposing a scalable and user-friendly solution for detecting manipulated media. By integrating advanced algorithms and real-world testing, we aim to contribute to the ongoing efforts to protect digital authenticity in an era of rapidly evolving AI technologies.",Cybersecurity,2024 – 2025,"Deepfake Detection, Deep Learning, Digital Media Integrity, AI Security, Image and Video Analysis",111
+P103,Fileless Malware Detection Using Deep Learning Techniques,"With the emergence of new fileless malware attacks that cannot be detected by conventional methods of anti-malware detection techniques, the cost of such attacks increases every day. The concept behind creating fileless malware attacks allows them to bypass traditional methods of detection and blocking. Fileless malware exploits legitimate system or user processes while residing inside the memory. The continuous evolution of those attacks requires countermeasures that can adapt to rapid changes. Utilizing artificial intelligence (AI) technologies to detect fileless malware attacks creates security controls that continuously evolve to keep pace with the ever-changing nature of cyberattacks. This project aims to develop a desktop tool that uses artificial intelligence to detect fileless malware through analyzing memory snapshots. The main target is to create a deep learning model that detects malicious instructions residing in the device memory. The tool blocks the infected process and alerts the user to suspicious activity. Detection accuracy is compared against several machine learning algorithms to ensure tool efficiency. Other related metrics, such as the detection time and the ratio of utilizing system resources, are also considered in the performance analysis. The proposed detection tool represents a contribution to current applied solutions to counter modern cyber threats and participate in protecting the digital transformation process and securing the community.",Cybersecurity,2024 – 2025,Fileless Malware; Malware Detection; Artificial Intelligence; Deep Learning.,112