UTAR Institutional Repository (Universiti Tunku Abdul Rahman)
Not a member yet
6132 research outputs found
Sort by
A study on the impact of the digital economy on China's manufacturing upgrading
No full textChina has long held the position of the “world's factory,” however, its manufacturing sector remains concentrated at the lower end of the global value chain. This structural limitation, combined with rising domestic and international challenges, has made manufacturing upgrading an urgent national priority. In this context, the digital economy has emerged as a key driver in advancing manufacturing upgrading. However, existing studies often lack comprehensive measurement systems and fail to systematically examine the mechanisms linking digital economy development to manufacturing upgrading. This study aims to examine the impact of digital economy development on manufacturing upgrading in China, investigate regional heterogeneity in this relationship, and identify the mediating roles of technological innovation, human capital, and transaction costs. A multidimensional framework is developed to measure manufacturing upgrading, comprising three key dimensions, namely manufacturing structural upgrading (MUS), manufacturing efficiency upgrading (MUE) and manufacturing environmental upgrading (MUC). The digital economy is measured using an indicator system comprising four dimensions and seventeen indicators, constructed through literature review and content analysis. The system Generalised Method of Moments (GMM) model is employed to test the impact of digital economy development on MUS, MUE and MUC. Furthermore, this research conducts the heterogeneity analysis and mediating effect analysis based on the GMM model. Results show that digital economy development significantly promotes MUS and MUC but has a negative effect on MUE. The negative effect may be due to the adoption of digital technologies requires significant adjustments in manufacturing processes, leading to a transitional period where inefficiencies may arise. Digital technology may weaken labour-intensive manufacturing' s competitiveness which China relied on, hindering labour productivity gains. Regional analysis reveals that all three dimensions are positively influenced by digital economy development in the eastern region, while the effects are mixed in the central and western regions. Mediation analysis indicates that technological innovation, human capital, and transaction cost mediate the relationship between digital economy and both MUS and MUE, while only transaction cost mediates the effect on MUC. The findings highlight the need for region-specific policies to enhance digital infrastructure, human capital, and institutional efficiency to fully harness the benefits of digital economy development in manufacturing upgrading. Further, policies should be emphasised to promote the positive impact of DE on MUS and MUC through strengthening technological innovation, enhancing human capital and reducing transaction costs.
Keywords: Manufacturing upgrading, digital economy, indicator system, dynamic panel model, mediating effect
Subject Area: HD9720-9975 Manufacturing industrie
"She killed someone? good for her!"": character and narrative analysis of women in the ‘good for her’ horror subgenre films
Get PDFThis research analyses the depiction of women in modern horror movies under the "Good for Her" horror subgenre. This horror subgenre focuses on female protagonists asserting autonomy and agency in oppressive environments in morally ambiguous manners. This research conducts a qualitative content analysis of five films: Luca Guadagnino's Suspiria (2018), Matt Bettinelli-Olpin and Tyler Gillett's Ready or Not (2019), Ari Aster's Midsommar (2019), Robert Eggers' The VVitch (2015), and Jordan Peele's Nope (2022). Using Feminist Film Theory as the analytical framework, it utilises character and narrative coding to examine the depiction of female characters. This approach provides a detailed exploration of how these films portray female agency, autonomy, and empowerment. The Coding Sheet for Character and Narrative Analysis has five sections under Character Coding (Agency and Autonomy, Expression of Rage and Anger, Empowerment through Violence and Self Defence, Rejection of Male Control and Influence, and Character Complexity and Depth) and five sections under Narrative Coding (Subversion of Horror Genre Tropes, Theme of Female Rage and Liberation, Representation of Female Solidarity or Isolation, Use of Symbolism and Visual Representation, and Narrative Arc of Empowerment).
The study addresses two key research questions:
1. How does the “Good for Her” horror subgenre depict female characters’ agency, autonomy, and expressions of rage through character behaviours and narrative structures?
2. How do female protagonists in this subgenre exemplify or defy the central tenets of Feminist Film Theory?
This research aims to contribute to the discourse on feminist horror cinema by highlighting the importance of the "Good for Her" subgenre in redefining representations of women in horror. It aims to emphasise these films not only reflect shifting cultural attitudes towards gender but also serve as a powerful medium for feminist storytelling.
This study finds that the “Good for Her” subgenre in modern horror reclaims female rage, autonomy, and power through violent or supernatural narrative arcs that challenge patriarchal structures and align with key tenets of feminist film theory.
Keywords: Feminist Film Theory, Good for Her Horror, Horror Films, Female Agency, Modern Horror Cinema, Patriarchy, Autonomy, Female Rage
Subject Area: PN1995.9.W6 – Women in Motion Pictures
Subject Area: PN1995.9.H6 – Horror Films (Genre Studies)
Subject Area: HM851 – Feminism / Feminist Theory
Subject Area: HM831 – Social Change
Subject Area: HM1001 – Social Psychology
Predictive personalized workout and dietary guidance system
Get PDFThe COVID-19 pandemic has increased the use of fitness and dietary mobile applications.
While existing fitness and dietary applications offer useful functionalities, they often fail to
deliver personalised recommendation that account for individual differences. This project
proposes the development of the “Predictive Personalised Workout and Dietary Guidance
System”, a comprehensive mobile application designed to address the shortcomings of existing
systems. This application utilises publicly available datasets and integrates artificial
intelligence to analyse user data such as weight, height, gender, and age, offering tailored
recommendations that evolve with user progress. Deep learning models were integrated and
evaluated to predict users’ Body Mass Index (BMI) classification. During the development
phase in, three predictive models were implemented and evaluated: a Deep Neural Network
(DNN), a U-Net-based Convolutional Neural Network (CNN) and a Random Forest. Among
them, the CNN model achieved the highest test accuracy of 90.36%, but DNN and Random
Forest only achieved a test accuracy of 88.70% and 85.00%, respectively, proving the U-Netbased CNN model is more effective and reliable for BMI classification. This result highlights
the advantage of using a U-Net-based CNN architecture for personalised health predictions
within the application. Unlike existing systems, which often focus primarily on exercise
tracking with minimal dietary support and lack of suitable workout recommendations. The
application will include functions such as fitness and dietary tracking, community platform to
enhance user engagement and motivation, artificial intelligence (AI) chatbot that support users
with personalised guidance, and weight tracking function. In addition, the system implements
a personalised dietary module that uses AI to analyse meals’ macronutrient intake, enabling
users to adopt a more health-oriented diet. With the comprehensive functions offered by the
system, users are expected to benefit from more precise and adaptable health guidance, thereby
improving long-term commitment and overall health outcomes
A personal financial management application with spending monitoring
Get PDFThis project introduces an advanced personal financial management application with
spending monitoring capabilities by tracking and monitoring their financial activities.
The field of study for this project encompasses financial technology (fintech), mobile
application development, and personal finance management. There are services such as
effective budgeting, expense tracking, and financial decision-making tools provided to
users by using mobile platforms.
Nowadays, many people facing some challenges to keep track of their daily expenses
due to busy schedules. Additionally, numerous payment methods have been introduced
in the market, including e-wallet platforms such as TNG eWallet, GrabPay, ShopeePay,
and embedded wallets within some mobile applications to ease customer payments.
However, this variety has further complicated expense tracking, as individuals must
keep track of multiple transactions across different platforms and navigate several apps
to review their spending history. Consequently, the process of documenting daily
expenses has become more difficult and time-consuming. To address this, there are
various expense recording applications have appeared in the market, offering features
such as receipt or transaction scanning through Optical Character Recognition (OCR).
However, many expense tracking apps still fall short in delivering an app that meet the
user requirement. To illustrate, the lack of automatic categorization of expenses into
different categories accurately and unable to extract the correct amount of expenses.
This project able to solve these limitations, by automatically categorizing expenses and
allowing users to correct the result, which enhance the flexibility. Further, some
rewarding system had been applied in this mobile application which aim to motivate
the users to manage their expenses wisely. Additionally, AI-driven savings tips have
been introduced to provide users with personalized financial guidance. The main
objective of this project is to offer customers a user-friendly interface that simplifies
their daily lives and reduce the effort required from users to record their expenses. Users
will be able to check their expenses or financial reports anytime such as daily and
monthly, helping them achieve their financial goals as well as making better financial
planning decisions
Developing a deep learning model to detect social media hate speech texts
Get PDFHate speech detection on online social media (OSM) platforms remains a significant
challenge due to the complexity of linguistic expression and inherent class imbalance
in available datasets. This study proposes a hybrid deep learning framework that
integrates DistilBERT with CNN and BiLSTM layers to perform multi-class
classification, categorizing input text into hate speech, offensive language, or neutral
classes. Five experimental configurations were conducted using the Davidson dataset,
including baseline training, resampling, class weighting, combined imbalance
mitigation, and an ablation study on preprocessing. DistilBERT was selected as the core
architecture to balance computational efficiency with representational power. The
baseline model achieved strong performance with 88.32% accuracy, a Cohen’s Kappa
of 0.6805, and a macro-AUC of 0.9260. Class imbalance mitigation techniques
demonstrated trade-offs: resampling and class weighting improved the minority class
F1-score to above 0.40 but reduced overall accuracy. The ablation study confirmed the
critical role of preprocessing, as the exclusion of data cleaning and tokenization
degraded minority class F1-score to 0.2886 despite stable accuracy. Overall, results
highlight the effectiveness of lightweight BERT-based architectures and produced
comparable result in detecting underrepresented class of data for hate speech detection
while emphasizing the importance of preprocessing and balanced training strategies for
equitable classification performance
Pet health and management system
Get PDFIn today’s fast-paced world, managing pet health has become increasingly challenging
due to the lack of comprehensive tools for monitoring and maintaining pet health. This
project introduces the concept of a pet health and management mobile application that
provides a holistic solution for managing pet health. The application enables pet owners
to easily track their pet’s health records, monitor symptoms, and get actionable advice
to ensure their pets’ health and vitality. For the user interface, the application allows
pet owners to effectively record important health data such as vaccination schedules,
medication history, and doctor visits. It also includes diagnostic features to help detect
early signs of disease based on observable symptoms with an AI-powered diagnosis.
Through personalized educational content, the application provides the AI-chatbot
assistance for educational purpose which enabling pet owners to gain the knowledge
needed to make informed decisions about their pet’s health. The application further
supports emergency situations by integrating geolocation capabilities, enabling users to
quickly find a nearby veterinary clinic or vet hospital based on their current location.
With its continuous learning capabilities, the application can adapt to the changing
health needs of pets and their owners. The goal of the Pet Health & Management mobile
app is to enhance pet care by providing pet owners with the tools to make informed,
proactive decisions and foster a healthier, more sustainable lifestyle for their pets. The
app addresses the growing need for user-friendly, efficient and technology-driven
solutions in the pet health management space
Development of an automated sustainable farm management system
Get PDFThis project presents the development of an automated sustainable farm management system aimed at improving agricultural efficiency through smart technology and eco-friendly practices. Farmers, especially small to medium-scale operators, often face challenges such as inefficient water usage, limited monitoring capabilities, and high dependency on manual operations. To address these issues, this project integrates Internet of Things (IoT) components including soil humidity sensors, automated water systems, and solar-powered operations to monitor and control key farming processes.
A mobile application is developed to enable real-time monitoring and remote control of watering schedules, lighting systems, and sensor data collection. Additionally, rainwater harvesting mechanisms and solar panels are incorporated to promote sustainability and reduce dependency on external resources. The system is developed using the Agile methodology, allowing for iterative testing and user feedback throughout the development cycle.
By automating repetitive tasks and promoting efficient resource utilization, the proposed system enhances farm productivity, reduces operational costs, and supports environmental sustainability. The outcome of this project provides a viable solution for modernizing traditional farming practices and contributes to the advancement of smart agriculture
The effect of personality traits and self-efficacy on academic performance among sutdents at a private university in Malaysia
Get PDFThe aim of this research is to study the effect of personality traits and self-efficacy on academic performance among students at a private university in Malaysia. Personality traits and self-efficacy may affect academic performance through openness to experience, conscientiousness, extraversion, agreeableness, emotional stability and self-efficacy among students. The researchers targeted undergraduate students from University Tunku Abdul Rahman (UTAR). Questionnaire has been distributed through Google Form. The researchers have successfully collected 377 responses. Statistical Package for Social Sciences (SPSS) Version 29 has been used to analyze and interpret the data collected for pilot study and full study. To test the significant relationship between the independent variables (openness to experience, conscientiousness, extraversion, agreeableness, emotional stability and self-efficacy), and dependent variable (academic performance), the researchers have used the Pearson Correlation Analysis, and Multiple Linear Regression Analysis. It has been found that all the independent variables (openness to experience, conscientiousness, extraversion, agreeableness, emotional stability and selfefficacy) have significant and positive relationships with the dependent variable (academic performance). The summary of major findings, implications of the study, limitations and recommendations of this study have been provided in this research. Keywords: Academic Performance, Openness To Experience, Conscientiousness, Extraversion, Agreeableness, Emotional Stability, Self-Efficacy Subject Area: BF698-698.9 Personalit
Rapid learning tools for student
Get PDFDue to information overload, language difficulties, and fragmented technologies, students frequently struggle to navigate large volumes of academic content. The creation of a web-based Rapid Learning Tool (RLT) that combines document analysis, AI-powered summarization, translation, definition lookup and research support into a single platform is the project’s suggested solution to these problems. A variety of document formats (PDF, DOC, DOCX, PPT, PPTX, and TXT) can be uploaded to the system, which also automatically extracts and summarizes important information, translates or explains words in real time, and collects relevant research papers from reliable databases, for example Semantic Scholar. Additionally, the word cloud visualization and keyword extraction features enable students to rapidly understand the primary ideas of texts. Rapid Learning Tool (RLT) combines AI services like Cohere AI, Ollama, DeepL, and Google Translate and is based in Node.js and Express.js for the backend. To guarantee a customized and safe educational environment, it also makes use of secure JWT authentication, session monitoring, and user conversation history abilities. Rapid Learning Tool successfully lowers students’ cognitive load while improving their understanding and learning effectiveness by using these features. It overcomes the gap between disjointed learning processes and effective, intelligent learning support by giving students a reliable, adaptable and useful instructional tool
The relationship between non-attachment, compassion, and prosocial behaviour among undergraduate students in Malaysia
Get PDFThe present study’s aim is to explore the correlation between non-attachment, compassion and prosocial behaviour and how these variables influence each other. The current study proposed three hypotheses suggesting that positive relationships exist between non-attachment, compassion and prosocial behaviour among undergraduate students in Malaysia, with the study’s findings indicating a significant and positive relationship between non-attachment and prosocial behaviour, compassion and prosocial behaviour, as well as non-attachment and compassion. This study is a cross-sectional and quantitative research. The aggregate number of participants recruited was 162 via purposive sampling method and their responses were accumulated through the internet questionnaire. All participants were Malaysian, undergraduate students from local universities aged 19–20 and most of them were Chinese. The present study included three instruments which were Non-attachment Scale–Short Form (NAS-SF), Compassion scale (CS), and Prosocialness Scale for Adults (PSA). Additionally, the relationships among these three variables was examined by using Pearson’s Correlation. The present study’s findings suggest that there is an expanded concept of virtue-meditation-wisdom and have addressed the gap, as there have been limited studies conducted to examine these three variables. Besides that, these findings also suggest that it can be embedded in the educational programs, counselling services and national youth development policy. In short, empirical support is provided by the present study on the relationships between non-attachment, compassion and prosocial behaviour