UTAR Institutional Repository (Universiti Tunku Abdul Rahman)
Not a member yet
6132 research outputs found
Sort by
Smart parking system with real-time parking lot status monitoring using Internet of Things (IoT) and Radio Frequency Identification (RFID)
Get PDFThe ever-growing number of vehicles in urban areas has significantly intensified the challenge of finding available parking spaces, leading to increased frustration for drivers and contributing to environmental pollution due to prolonged vehicle idling and unnecessary driving. Real-time updates and efficient space utilization are rare features of most parking management systems on the market today. Apart from this, visibility is usually limited, especially in large parking lots, and the signs can be unclear or difficult to read, which causes drivers to have no idea where the available parking spots are. This project presents a Smart Parking System using IoT and RFID Technology designed to address these issues by providing a real-time parking lot status monitoring solution. The system integrates various hardware components, including eight IR sensors to detect vehicle presence at the parking slots, two servo motors for gate control, two ESP32 microcontrollers, two RFID readers with tags, an LCD display for showing parking status and time, an OLED display for showing RFID scanning messages, and a buzzer for audio feedback. Additionally, two IR sensors monitor vehicle presence at each of the gates. Besides that, custom developed HTML-PHP integrated web pages enable public users to view the status of each parking slot in realtime, whether it is “AVAILABLE”, “OCCUPIED”, or “RESERVED” and then access features like sign-up, login, top-up, and reservation, which allow the users to reserve a parking slot for a particular time before arriving. The user information, RFID scanning timestamps, reservation user and time, and status of each parking slot are stored in a MySQL database. Users can obtain parking information on the web page using their mobile devices by scanning a QR code or by visiting the provided URL link and logging into their registered accounts. The system operates by detecting vehicles at the entrance, verifying RFID UID with database, and managing gate operations based on slot availability and account balance. Furthermore, a ESP32-CAM is used to capture vehicle images at the gate entrance when the IR sensor detects movement and save them to Google Drive. The performance of the system will then be assessed based on its accuracy. Based on the results, the system prototype achieved 100 % accuracy. When compared to other proposed Smart Parking Systems, it is clear that this method is more cost-effective and reliable. This system aims to reduce parking-related frustrations and environmental impact to enhance overall urban mobility and sustainability.
Keywords: IoT, RFID, Smart Parking System, Real-time Monitoring, Reservatio
Medibot: UTAR hospital AI health companion
Get PDFThis proposal introduces a project aimed at enhancing the user experience on UTAR Hospital’s platform by developing an English-Chinese multilingual chatbot that provides personalized medical guidance through doctor recommendations and disease prediction. The chatbot leverages advanced technologies such as the LLaMA transformer model, Retrieval-Augmented Generation (RAG), and natural language processing (NLP) to interact with users in a natural, friendly, and informative way.
The core of the project lies in the chatbot’s ability to understand user-described symptoms and predict the most likely disease category using machine learning techniques, such as Random Forest Classifier, Logistic Regression, Xgboost Classifier. Based on the prediction, the chatbot recommends suitable doctors from UTAR Hospital’s Traditional and Complementary Medicine Centre for further consultation.
RAG plays a key role in generating human-like responses by combining retrieved information with natural language generation, ensuring the conversation feels more engaging and helpful. The chatbot’s multilingual capability, supporting both English and Chinese, enables it to assist a wider and more diverse range of users, particularly in Malaysia’s multicultural context.
Additionally, the system incorporates a similarity search mechanism using a temporary vector database to improve the accuracy and relevance of responses. It also features an integrated online appointment system to streamline consultation scheduling and reduce reliance on manual processes.
Overall, this project aims to enhance healthcare accessibility through a multilingual chatbot that supports a diverse user base by providing symptom-based disease prediction, personalized doctor recommendations, and streamlined appointment scheduling based on the predicted disease category
Cleadr : AI-enhanced AR navigation app for seamless driving
Get PDFNavigation systems have become an essential tool for drivers to navigate through
journeys with ease and confidence. However, navigation systems are far from perfect
as they still pose some limitations such as ambiguous directions, insufficient real-time
assistance, and poor User Experience (UX). These limitations of navigation systems
often lead to confusion and uncertainty for the driver. Hence, this project proposed a
solution by integrating Augmented Reality (AR) and Artificial Intelligence (AI) into a
navigation system. Specifically, a mobile AR navigation application integrated with AI
assistance was proposed. The development of the system adopted the agile Extreme
Programming (XP) methodology that allowed for quick and iterative development, as
well as ample flexibility in responding to changing requirements. Core technologies
involved were Flutter, Unity, PyTorch, and TensorFlow. The development of the
system was structured into 3 core modules: the Maps Module, Navigation Module, and
Intelligence Module. Core features of the system included the AR navigation and lane
identification. AR navigation provided clearer directions by projecting them onto the
real-world environment, while lane identification provided context-awareness for
effective lane change instructions. A system performance evaluation was conducted
with performance metrics such as response time and accuracy. The system was
responsive with an overall response time of 1.80 seconds. Additionally, a lane
identification model was trained from a custom Malaysian highway roads dataset that
consisted of a total 22,806 images. The model managed to obtain an accuracy of
99.21%. In summary, this project successfully developed a mobile AR navigation
application integrated with AI assistance to achieve the outlined project objectives.
Moreover, the project contributed a lane identification model accustomed for
Malaysian highway roads with reliable accuracy
Smart financial tracking mobile application
Get PDFIn this project, a Smart Financial Tracking Mobile Application is proposed to cater to all individuals who care about being out of debt, spending, budgeting, and financial tracking in Malaysia. Financial tracking, also known as expense tracking, is a common approach whereby an individual manages their expenses by recording their daily, monthly, and yearly expenditure through digital software such as Microsoft Excel, cross-platform budget tracking applications like the popular You Need A Budget (YNAB) or through traditional financial entries on notebooks. Most financial tracking systems have limitations, such as a lack of Asian Bank Integration, insufficient financial data insight, and mundane financial entry. The proposed system solves the common mobility issue for users who want to track their finances on the go and aims to solve the problems mentioned earlier. Moreover, it also leverages state-of-the-art AI technology, such as a seamless Integration with Google’s newest Machine Learning Kit models for near real-time receipt extraction and various connections with third-party APIs such as LangChain API, to improve data extraction accuracy. The core features of the mobile application include receipt scanning with OCR, scraping email financial data, chatting with financial data leveraging Large Language Models (LLM) like OpenAI’s ChatGPT, Malaysian bank app integration (Maybank, CIMB, Public Bank) and voice data recognition entry. Furthermore, the main Software Development Life Cycle (SDLC) model used in this project is Rapid Application Development (RAD). This approach enables quick creation of multiple prototype versions that can be refined based on user feedback. Lastly, the main tools used for development are IDE, such as Android Studio and Visual Studio Code, Firebase as the backend as a service, a mobile phone, and a laptop.
Area of Study: Mobile Application Development, Artificial Intelligence
Keywords: Financial Technology integration, Optical Character Recognition, Voice Recognition, Expense Tracking, Generative AI, Workflow Automatio
Risk factors, coping strategies, and effects of intimate partner violence: Triangulation of scoping review and interviews with survivors and social workers
Get PDFIntimate partner violence (IPV) remains a critical global health issue, with WHO reporting 30% prevalence among women worldwide. In Malaysia, IPV rates are increasing, yet research remains limited. This study addressed this gap through a multi-method investigation combining scoping reviews (Study One), interviews with IPV survivors (Study Two), and social worker perspectives (Study Three), with findings triangulated in Study Four. In Study Four, the results of these three studies were then triangulated to identify findings that are robust, possible and less likely. The risk factors reported in the scoping review and interviews were classified using the Ecological Framework, pointing towards the fact that risk factors exist across all four levels of the framework and interact to predict IPV among women. Coping strategies reported in the scoping review and interview responses were categorized according to Skinner et al. (2003)’s 11 families of coping, providing further backing for the framework’s applicability in IPV research. Finally, the effects of IPV that were reported in the scoping reviews and interviews were classified according to the Biopsychosocial Model, emphasizing the profound impacts of IPV, with psychological effects—especially mental health issues—being the most reported effect. In Study Four, the triangulation of results from Studies One to Three identified robust and possible risk factors, prevalent and possible coping strategies, as well as robust and possible effects of IPV. These findings highlight IPV's multidimensional nature, demonstrating how risk factors interact across ecological levels while emphasizing the predominance of psychological impacts. The study underscores the necessity for comprehensive, multi-level interventions addressing prevention, survivor support, and policy reform. Results significantly advance Malaysia's IPV research landscape while providing an empirical foundation for developing culturally-appropriate interventions and guiding future investigations into this critical public health issue
Transitivity analysis on novels used in Malaysian secondary schools for Sijil Pelajaran Malaysia (SPM) English literature subject
Get PDFWith the indeterminatenature of language use shaped by individual worldviews and experiences,this studyinvestigated the transitivity choicesliterary writers make to convey experiential meaning.Focusing on the significance of literary texts in society, it addressed challenges faced by writers in expressing local thoughts and realities, linguistic complexities and ambiguity, the use of language to convey social issuesandideologies, and the limited studies on literary texts through the lens of transitivity system. UsingHalliday andMatthiessen’s (2014) Transitivity System, this study analysed The Clay Marbleand The Lost King, two novels studiedin Malaysian secondary schools for the SPM English Literature subject. In The Clay Marble, relational process (64) was the most frequently used, followed by material process (38) and mental process (31). InThe Lost King, relational process (56) was similarly dominant,followed by material (27) and mental processes(24). Employing textual analysiswith transitivity analysis as its means, this study identifiedthemesand provided insightsintothe representationsof the characters and events. The findings indicatedthat transitivity choices played a crucial role in shapingthe representation of characters and events as different transitivity choicesconvey distinct meanings, ideologies andthemes. Furthermore, transitivity analysis revealed the social issues embedded in the novels, offering a deeper understanding of how language choices reflectsocial realities.In conclusion, this study demonstrated that transitivity analysis is an effective methodological tool for uncovering meaning in literary texts, revealing how language functions as a representation of human interactions and societal contexts
The moderating effect of self esteem on the relationship between social connectedness and teaching anxiety among teachers in Malaysia
Get PDFThis study aims to investigate the moderating effect of self-esteem on the relationship between social connectedness and teaching anxiety among teachers. Most of the past studies are less likely to select teachers to investigate these variables. It is crucial to examine teachers’ mental health as teachers are found to have high teaching anxiety levels in Malaysia. With a low level of social connectedness, teachers may face high teaching anxiety levels when handling their teaching tasks as they are unable to seek psychological support from others. Self-esteem as a teacher might serve as a powerful moderator to influence the relationship where it can directly influence the confidence and passion for teaching among teachers. It is conducted with 390 teachers in Malaysia. The sample is collected by purposive sampling method to reach the teachers who fulfill the sampling element heterogeneous based on the types of school, school location, teachers’ gender and year of working experience. The questionnaires include the Psychological Sense of School Membership, Teaching Anxiety Scale and Rosenberg Self-Esteem Scale via online distribution. Pearson Correlation is employed to assess the association between social connectedness, teaching anxiety and self-esteem while Hayes SPSS Macro Process Model 1 is adopted to examine the moderating effect of selfesteem on the relationship between social connectedness and teaching anxiety. This research eventually comes up with a main outcome where self-esteem has a strong moderating effect on the relationship between social connectedness and teaching anxiety. With this finding, it has contributed to the literature pool in few aspects and also provides valuable insight for schools, policymakers and authorities who can make changes regarding teachers’ psychological health issues. The main contribution of this study is that it can provide research evidence and support to the future researchers who would need to implement related training plans related to self-esteem to the teachers
Integrating natural language processing (NLP) for enhanced stock market prediction trhough text and news data fusion
Get PDFThis study addresses critical challenges in stock market forecasting by introducing
FusionStockBERT, a transformer-based, multi-modal model that predicts both nextminute
price movement direction (up/down) and expected return — the latter
reframing traditional stock price regression into a more stable return regression task.
Existing NLP methods in stock market prediction faced limitations from the context
drift, extrapolation errors, out-of-vocab issue in lexicon, growth issue of dictionary
size and the inability to capture complex textual semantics when being applied to
the stock market movement prediction task. To address these issues, we proposed a
self-supervised learning to further fine-tune the FinBERT (a pre-trained BERT
model) directly on directional movement labels and augment its final [CLS]
representation with engineered trading features via an intermediate neural fusion
layer for downstream tasks.
Evaluated on minute-level Bloomberg news transcripts paired with trading data,
FusionStockBERT achieves 80.53% accuracy on the development set and 71.42%
on a held-out validation set for directional movement prediction—substantially
outperforming both non-attention CNN baselines and majority-class baselines. In
the return regression task, it delivers an MSE of 8.9×10⁻⁴ on the validation set,
demonstrating competitive precision in estimating the directional movement’s return.
These results highlight that integrating fine-tuned transformer embeddings with
structured market data provides a powerful, real-time tool for high-frequency trading
decision support
Focusguard: Real-time self-monitoring system for enhancing student focus using computer vision
Get PDFA real-time computer vision tool called FocusGuard aids students in focusing and avoiding distractions while studying. The system uses Eye Aspect Ratio (EAR) to detect drowsiness, head pose estimation to monitor attention, and Mouth Aspect Ratio (MAR) to detect yawns to prevent fatigue. These three crucial components of student focus are addressed by the system. The solution encourages organized study sessions by combining these elements with a Pomodoro Timer. Key facial features are tracked by the system using MediaPipe's facial landmark detection. FocusGuard gives students timely audio alerts to help them focus again when it detects signs of distraction or drowsiness. An early warning sign of possible fatigue is the yawn detection, while the head pose analysis detects whether students are avoiding their eyes from their study materials. The combination of these monitoring features with the Pomodoro Technique, which divides study sessions into concentrated work periods and selected breaks, is a significant innovation. This combination results in a clever study partner that keeps an eye on and directs student conduct. Students can monitor their focus patterns over time with the system's web-based interface, which shows session statistics and real-time metrics. The system's ability to detect periods of fatigue and diminished attention is demonstrated by preliminary testing. FocusGuard is a real-world example of how computer vision technology can be used to enhance student productivity and learning
Tech-savvy investors - Adoption of wealth tech for wealth management in Malaysia
Get PDFThe rapid advancement of financial technology has revolutionised the wealth management industry, leading to the rise of digital-only banks. This research examines the key factors influencing the adoption of Wealth Technology, including performance expectancy, effort expectancy, social influence, facilitating conditions and technology readiness. The research integrates the Unified Theory of Acceptance and Use of Technology (UTAUT) and the Technology Readiness Index (TRI) to develop a comprehensive theoretical framework. A quantitative research approach was employed, utilizing survey-based primary data collection from 384 Malaysians across different genders, age groups, highest educational levels, races, employment status, net income levels and regions. Statistical analysis was conducted to determine the significance of these factors in shaping user adoption behavior. The findings reveal that performance expectancy and effort expectancy, while social influence and facilitating conditions also play a significant role. Additionally, technology readiness contributes to user behavioural intention to adopt Wealth Tech in Malaysia. These insights provide valuable implications for financial institutions aiming to enhance user trust and engagement in digital banking services. This research contributes to the growing literature on fintech adoption and offers strategic insights for financial institutions and policymakers to enhance digital wealth management services.
Keywords: Digital Wealth Management, Wealth Tech, Technology Adoption, Behavioural Intention, Unified Theory of Acceptance and Use of Technology (UTAUT), Technology Readiness Index (TRI), Performance Expectancy, Effort Expectancy, Social Influence, Facilitating Condition, Technology Readines