UTAR Institutional Repository (Universiti Tunku Abdul Rahman)
Not a member yet
6132 research outputs found
Sort by
Application of genetic algorithm and JFugue in an evolutionary music generator
Get PDFThis project explores the application of Genetic Algorithms (GA) with JFugue, which is a Java-based music programming library to develop an Evolutionary Music Generator. The challenges with automating the music composition lies in capturing the complexity, creativity and emotional expressiveness of music which requires a deep understanding of musical theory. This project will resolve this problem by applying GAs to evolve musical compositions and offer a fresh view on computational music generation. It is relevant because it generalizes this process of music creation and provides instruments to musicians and non-musicians for diving into new, unique forms of musical expression. Genetic Algorithms will be chosen because they constantly improve solution, hence appropriate for the development of compositions consistent with the creative and harmonic principles. In addition, JFugue is an open-source, Java- based music programming toolkit that has been integrated to efficiently represent and manipulate the music, hence implementing the evolutionary process. This method proves to be effective by systematically exploring and generating a variety of musical ideas through iterative applications of selection, crossover, and mutation genetic operations. The appliance of harmonic principles, rhythmic patterns, and user feedback provides the basic for fitness assessment in ensuring that the generated music meets pre-set requirements. Music that has been generated using JFugue involves real-time generation and user-driven evolution. This will involve the explanation of the use of evolution algorithms combined with the music programming to be able to create creative digital music
Smart student timetable planner
Get PDFTimetable planning is a crucial yet challenging task for university students, as traditional manual methods are often time-consuming, prone to errors, and lack collaborative support. Students frequently face difficulties in avoiding timetable clashes, managing personal preferences, and coordinating with peers, which can lead to inefficiencies and added stress. To address these issues, this project introduces the Smart Student Timetable Planner, a system developed to streamline academic scheduling by providing both automated and manual timetable management options. The objectives of this project are to generate conflict-free and customizable schedules, enable real-time collaboration among students, and offer administrative tools for maintaining course information. The project scope encompasses features such as secure login, course selection with conflict detection, timetable history, comparison between auto-generated and manual schedules, collaboration modules, and export functionality. To achieve these objectives, the system adopts the Rapid Application Development (RAD) methodology, ensuring iterative design, prototyping, and user feedback integration throughout the process. The system is implemented using Node.js with Express for server-side development, HTML, CSS, and JavaScript for the frontend, and Socket.IO for real-time collaboration. Course data is managed in CSV format, parsed into JSON for fast processing, while sessionStorage and localStorage handle user data within active sessions. A Genetic Algorithm forms the core scheduling engine, generating optimized timetables that respect both hard constraints, such as avoiding clashes, and soft constraints, such as personal preferences.The final output of this project is a functional web-based timetable planner that successfully enhances scheduling efficiency, reduces the likelihood of errors, and improves the overall academic planning experience. With its flexible design, collaborative features, and administrative integration, the Smart Student Timetable Planner demonstrates significant potential as a scalable solution for modern university scheduling needs
Visionserve: A vision based food detection system for fast food restaurants
Get PDFThis project explores the development of a vision-based food detection system using advanced computer vision and image recognition techniques, optimized for use in fast-food restaurants in Malaysia. The goal of this system is to automate the detection of food items, thereby enhancing operational efficiency, minimizing human error, and improving overall customer satisfaction. The driving force behind the project is the real challenges faced daily in McDonald’s restaurants, where rapid and efficient service is critical. Customers often receive wrong items, which leads to frustration and long queues at the counter. Staff also spend time re-checking orders repeatedly, slowing down operations and increasing the risk of mistakes. During peak hours and promotional events, when restaurants are crowded, these issues become even more pressing, placing additional stress on staff and customers alike. Mistakes made in identification and distribution of food items can cause complaints from customers and potential loss of business for restaurateurs. This technology is thus presented as an adjunct technology solution, not as a replacement for humans, but as an intelligent tool that supports and supplements human force and enhances fast food restaurant’s universal service platform. The system utilizes computer vision methods trained to recognize various fast-food items such as burgers, fries, and drinks. When recognized, the system summarizes recognized food items on the tray. Deliverables under the project involve creating and launching an operational prototype, verifying its performance on the ground level on the terms of simulation within a fast-food restaurant condition, and total technical documentation for any potential commercial deployment or further works. Lastly, but significantly, this vision-based food detection system is one of the pacesetters of Malaysian fast-food operations in modernization. It improves service dependability, accommodates workforce diversity, and positions fast-food restaurant chain as a leader in embracing artificial intelligence and intelligent automation technology in the quick-service restaurant sector
Teaching proficiency through reading and storytelling (TPRS) as a technique to improve vocabulary skills of Malaysian secondary school students
Get PDFThis study explores the effectiveness of Teaching Proficiency through Reading and Storytelling (TPRS) as a technique to improve vocabulary skills of Malaysian Secondary School Students. The objectives of the study are to investigate the effectiveness of using TPRS to improve secondary school students’ English vocabulary skills and identify the students’ responses towards the TPRS method. A group of 47 students participated in a four-week program where vocabulary was taught through TPRS method. Pre- and post-tests were conducted to measure the effectiveness of the TPRS method, and the students’ responses were collected through a questionnaire. The pre- and post-test data were analyzed using Microsoft Excel and t-tests while the questionnaire was analyzed using descriptive method and past studies. The findings revealed that there are significant improvements in vocabulary skills after the intervention. The students also demonstrate positive attitude toward the TPRS method. According to the findings, this study suggested that the TPRS method can be an effective approach when it comes to language learning and apply in similar classroom settings to support language learning
AI-powered CCTV for intuder and visitor detection with IoT alert system
Get PDFAI-powered technologies that offer real-time detection and response are slowly replacing passive CCTV in surveillance systems. Traditional systems require constant monitoring and lack instant alert mechanisms, creating gaps in accessibility and responsiveness. This project addresses these limitations by developing a smart CCTV system that integrates artificial intelligence with IoT for improved security monitoring. The main objective is to design and implement an AI-powered CCTV capable of detecting intruders and visitors in real-time while providing instant mobile alerts and door control. The system was implemented using Python, OpenCV, Dlib for face recognition, YOLOv8n for human detection, and Telegram Bot API for IoT-based alerts. An Arduino-controlled servo motor was added to simulate automated door access for authorized users. Motion detection with MOG2 background subtraction triggered further analysis to balance efficiency and accuracy. Experimental results showed reliable performance across modules. Motion detection achieved high accuracy under normal conditions, face recognition reached up to 90% accuracy, and YOLOv8n human detection recorded an F1 score of 0.922 on test data. Telegram alerts were delivered within 1–2 seconds, ensuring timely notifications. In conclusion, the developed prototype demonstrates that integrating AI and IoT can deliver an affordable and effective smart surveillance system. While performance drops in lowlight conditions remain a limitation, the system provides a practical foundation for scalable home security with potential for future enhancements such as night vision integration, latency reduction, and expanded IoT features
《侠客行》中石破天的“自我技术”研究 : A Study of Shi Potian's ‘Technologies of the Self’ in The Ode to Gallantry
No full text不同于金庸其他武侠小说，《侠客行》摒弃了传统的家国叙事架构，在去历史化的纯粹江湖场域中，聚焦于一位纯真少年的成长历程。《侠客行》的创作不仅标志着金庸从宏大历史叙事向微观人性探索的艺术转型，更与二十世纪中后期西方兴起的后现代主义思潮形成跨时空的呼应。值得注意的是，彼时法国哲学家米歇尔·福柯（Michel Foucault）正经历着从权力谱系学向“自我技术”（Techniquede soi）研究的理论转向，二者不约而同地将思想旨趣聚焦于人的主体性问题。基于此，本文试以福柯的“自我技术”理论作为分析框架，引入中国传统儒释道思想作为补充参照，采用文献研究法、文本细读法与跨文化比较法相结合的方法论路径，系统分析《侠客行》中权力规训与主体建构的动态关系，旨在为金庸武侠研究提供新的理论视域。具体而言，本研究首先从微观权力视角切入，解析江湖社会中他者凝视、全景敞视等规训手段的权力运作机制。其次，立足于石破天这一具有突破性意义的文学形象，剖析其以纯真本性与直言勇气突破权力桎梏的审美化生存方式。最后，围绕“我是谁”的哲学命题，揭示主体如何通过创造性的生命实践，在规训体系中开辟自由之境，不断实现自我的重构与主体性的超越。【关键词】金庸、《侠客行》、石破天、米歇尔·福柯（Michel Foucault）、自我技术（Technique de soi
Smart mobile application for resume building and career advisory services
No full textThe increasing complexity for the current modern job market trend has clearly stated that the importance of professional resumes and career advisory services for job seekers no matter students, graduates, and professionals. By reviewing the current mobile resume building applications, it shows that these applications often lack AI functionality, integration of career advisory features, and sufficient customization options which limits user effectiveness. To address these issues, this project aims to develop a smart Android mobile application, Job Assistant, with integrates Google Gemini AI model for main features. The project follows Scrum methodology framework by conducting the requirement analysis, project designs, development, testing and deployment across five sprints. The developed application used Android Studio, React Native, Visual Studio Code, SQLite database, and integrated Gemini AI by using API key for supporting the professional resume creation and personalized career advice recommendations. The developed key features include resume creation with AI-powered, AI generation career advisory services and various customizable options for resumes. Testing with unit test and UAT gained an overall good result with overall unit test cases passed and UAT 4.82 over 5 rating. These findings highlight that integration of AI into resume-building tools not only increases efficiency but also strengthens user job opportunities with supporting career decisions. To conclude, this project, the smart mobile application with resume building and career advisory services which called Job Assistant represents a valuable community initiative software solution that supports equal access to career development resources and align with Sustainable Development Goals (SDG) such as decent work and no poverty. This Android mobile application can be further expanded with features like multi-languages support, integration with job platforms, iOS version release, offline AI functionality and others to maximize and improve this application.
Keywords: Resume Builder, Resumes, Career Advisory Services, Android Mobile Application, Software Development, React Native, AI, Job Opportunity.
Subject Area: QA7
The influence of maturity to parenthood on fertility intention among Malaysian childless married couples: Examining the moderating role of gender
Get PDFAs fertility rates continue to decline globally, understanding the psychological factors influencing individuals’ fertility intentions becomes increasingly important. This study aimed to examine how different dimensions of maturity to parenthood—valence, behavioural, and cognitive-emotional—predict fertility intention, and whether gender moderates these relationships. A quantitative, cross-sectional research design was employed. Participants were recruited through non-probability sampling methods, including purposive, self-selection, and snowball sampling, resulting in a sample of 95 married individuals, currently childless, aged between 20 and 44 years (M = 31.23, SD = 5.25). Most participants were female (n = 71), followed by male (n = 24). The study was conducted in Malaysia, and data were collected via online self-report surveys. Instruments included the Desire to Avoid Pregnancy Scale, assessing fertility intention and the three subscales of the Maturity to Parenthood Scale. Data were analysed using SPSS, including Pearson correlation analysis, hierarchical multiple regression analysis and Hayes’s PROCESS macro (Model 1). Results indicated that behavioural maturity significantly predicted fertility intention, whereas valence and cognitive-emotional maturity did not. Gender did not moderate the relationship between any maturity dimensions and fertility intention. These findings suggest that psychological maturity, particularly practical preparations, is more strongly related to fertility intention than valuing parenthood or holding realistic and emotionally grounded attitudes. No gender moderation may reflect shifting gender roles and shared practical and relational considerations, as well as similar viewpoints in family planning across genders. This study provides useful implications for fertility education, reproductive health interventions, and policymaking that promote earlier behavioural readiness toward parenthood
Secured agriculture sensor data based on end-to-end encryption using raspberry pi
Get PDFThe rise of smart agriculture in Malaysia, powered by IoT sensor networks, has transformed farming by enabling real-time monitoring of soil moisture, temperature, and environmental conditions. However, much of this sensor data is transmitted without encryption, exposing it to risks such as interception, tampering, and unauthorised access. This project addresses these security concerns by developing a secure data collection and visualisation system using Raspberry Pi and Node-RED. The system integrates multiple robust encryption algorithms—AES-128, AES-256, ChaCha20, and Twofish—for end-to-end data protection. Additionally, a benchmarking tool was developed to evaluate and compare the performance of these algorithms in terms of speed, memory, CPU usage, and encryption overhead. The final outcome is a lightweight, replicable solution for secure smart agriculture systems that enhances trust, integrity, and data privacy in IoT-based farming environments
Cryptanalysis of elliptic curve scalar multiplication algorithms
Get PDFThis project explores scalar multiplication algorithms in Elliptic Curve Cryptography, focusing on the binary method and elliptic net method applied in Elliptic Curve Diffie-Hellman and Elliptic Curve Digital Signature Algorithm. Scalar multiplication is the most computationally intensive operation in Elliptic Curve Cryptography and directly impacts both cryptographic strength and performance. There is lack of standardized scalar multiplication algorithm or parameter set to ensure compatibility and interoperability in cryptographic implementations. This creates challenges in developing secure Elliptic Curve Cryptography systems and performing cryptanalysis for scalar multiplication algorithms. This research implemented both methods on secure Twisted Edwards curves (numsp384t1 and numsp512t1) using the affine coordinate system for clearer point representation. The binary method uses a double-and-add approach, which introduces conditional branches that increase execution variability, making it more vulnerable to timing-based side-channel attacks. In contrast, the elliptic net method structures point operations more uniformly, reducing observable patterns and improving leakage resistance despite its higher complexity. Simulated attack scenarios, including timing and power analysis, revealed that the elliptic net method maintained more consistent behavior and offered better protection against information leakage. Overall, the findings highlighted the performance of Elliptic Curve Cryptography Scalar Multiplications over side-channel attacks in the implementations