UTM Press Journal Management (Universiti Teknologi Malaysia)
Not a member yet
7667 research outputs found
Sort by
Analisis Kecenderungan Penggunaan Percampuran Kod Bahasa Inggeris dalam Video Masakan di Tiktok: An Analysis of the Tendency to Use English Code-Mixing inCooking Videos on TikTok
No full textPercampuran kod sudah tidak asing lagi dalam setiap pertuturan majoriti masyarakat di Malaysia. Lazimnya, terdapat dua campuran bahasa yang sinonim digunakan iaitu, bahasa Melayu dan bahasa Inggeris. Fenomena ini semakin menjadi-jadi dan telah menjadi kebiasaan bagi segelintir masyarakat khususnya menerusi penggunaan media sosial di Tiktok. Tiktok merupakan sebuah platform yang menyajikan kandungan video berdurasi pendek yang lebih separuh penggunanya terdiri daripada golongan muda. Generasi muda beranggapan bahawa penggunaan percampuran kod tersebut membuatkan gaya komunikasi mereka terlihat lebih moden. Percampuran kod secara berlebihan telah mengganggu gugat martabat bahasa Melayu sebagai bahasa ibunda di negara ini. Penggunaan bahasa Barat seperti bahasa Inggeris ini dalam pertuturan lebih meluas diguna pakai dalam Tiktok. Oleh itu, kajian ini dilakukan untuk mengenal pasti percampuran kod dalam video masakan di Tiktok. Oleh hal yang demikian, pengkaji menganalisis percampuran kod dalam video masakan Tiktok berdasarkan pendekatan K. Kanthimathi (2007). Kajian kualitatif ini dijalankan menggunakan kaedah pemerhatian video dan juga analisis kandungan. Hasil kajian mendapati bahawa fenomena percampuran kod antara bahasa Melayu dan bahasa Inggeris banyak digunakan dalam video masakan di Tiktok. Kemudian, analisis berdasarkan pendekatan K.Kanthimathi (2007) menunjukkan penggunaan pelbagai bentuk percampuran kod dari segi fungsi sikap, fungsi ekspresi dan fungsi arahan
Social Support-Seeking among People with Mental Illness on Social Media: A Systematic Review
No full textWith the rise of social media, understanding how people with mental health illnesses seek social support online is crucial for enhancing support mechanisms. This systematic review synthesises evidence on the information-seeking behaviours of individuals with mental illness on social media, following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) framework. A thorough search was done for studies published between 2014 and 2024 on PubMed, ProQuest, Web of Science, and Google Scholar. Studies that looked at how people with mental health illness use social media to seek social support were required to meet inclusion criteria. A standardised form was used to extract the data, which were then qualitatively synthesised. A total of 11 studies met the inclusion criteria. The findings indicate that people with mental illness use social media primarily to seek social support, share personal experiences, and gather information on symptoms, treatments and medications. Social media platforms like Facebook, X, and Reddit were commonly used. This review highlights that social media is a significant resource for people with mental illness’ information and support seeking. While it offers benefits, there are also concerns about the reliability of information and privacy issues. Future research should focus on interventions to enhance the positive impacts of social media on social support-seeking behaviours
A Systematic Review on the Use of Artificial Intelligence in Providing Feedback in Teaching Writing
No full textArtificial Intelligence (AI) has rapidly evolved in recent years, permeating many facets of human life. One of the noteworthy uses of AI in education is the provision of feedback, which plays a crucial role in enhancing students’ learning outcomes, particularly in writing skills. Providing feedback on students’ writings is important for both teachers and students. Feedback in writing is a fundamental aspect of teaching practice, and AI offers innovative solutions to enhance its effectiveness. This review performs a rigorous systematic review of the existing literature on the utilization of AI in educational feedback mechanisms. Through a meticulous examination of scholarly literature, this review synthesizes findings regarding the effectiveness, challenges, and future directions of AI-driven feedback mechanisms. The findings reveal the potential of AI in providing personalized and timely constructive feedback. However, it also addresses challenges like the integration of AI tools into pedagogical practices to cater students from different educational levels. Additionally, it identifies the gaps in the existing literature and offers insights into potential areas for further exploration and development to optimize AI-driven feedback in improving pupils’ writing skills. To improve writing instruction, this review hopes to serve as a valuable resource for educators, researchers and policymakers seeking to leverage AI for feedback provision
EFFECT OF CENTRIFUGATION SPEED OF PINEAPPLE EXTRACT AS NATURAL COAGULANT ON RUBBER CHARACTERISTICS
No full textNatural rubber is one of the leading plantation commodities widely cultivated in Indonesia to produce products useful for human needs. This research aims to determine the effect of centrifugation speed on the properties of PB 260 clone rubber using pineapple extract coagulant. Among rubber farmers, there are obstacles in processing natural rubber, such as reliance on chemical coagulants and the lack of cheap and safe alternatives. Therefore, it is necessary to find coagulants that are affordable, safe, and effective. Pineapple extract was used as the natural coagulant in this research. The latex used was 150 mL of PB 260 clone rubber, with 75 mL of 100% natural coagulant and 2% of formic acid. The rubber properties tested included Initial Plasticity (PO), Plasticity Retention Index (PRI), Mooney viscosity, dirt content, ash content, volatile matter, Dry Rubber Content (DRC), and FTIR analysis. The research results showed that the highest values for PO, PRI, DRC, ash content, dirt content, and mooney viscosity values of 39.5%, 88.61%, 30.95%, 0.48%, 0.19%, and 62 MU respectively, were obtained using 0 rpm pineapple extract (without centrifugation), with. Meanwhile, volatile matter content was highest at 0.37% with pineapple extract centrifuged at 5000 rpm. This study revealed that higher centrifugation speeds increased the pH of the natural coagulant and decreased the concentration of H+ ion, which led to an accelerating of the latex coagulation process. So it can be concluded that pineapple extract can be used as a substitute for chemical coagulants as it effectively coagulates latex. The rubber from coagulation process using pineapple extract complies with Indonesian National Standard (SNI) for rubber characteristics. Centrifuged pineapple extract offers lower value of impurities and ash content while non centrifuged extract yields higher value of Po and PRI
NOVEL DIFFERENTIAL EVOLUTION FOR FEATURE SELECTION IN ANOMALY-BASED INTRUSION DETECTION
No full textIn recent years many organizations and end users suffer from cyber-attacks or intrusions also known as zero-day attacks that aim for damaging resources or theft of data. A well-known tool for detecting such intrusions is anomaly-based Intrusion Detection System (IDS). IDS have integrated Evolutionary Computation (EC) algorithms as dimensionality reduction method to enhance the detection performance. A major limitation in anomaly-based IDS is the high rate of false alarms due to several reasons most importantly is the high volume of training and testing datasets. These high dimensionality datasets could contain irrelevant, duplicate, and redundant features that cause misclassifications and increase the false alarm rate. In this research a new variant of Differential Evolution (DE) algorithm called Differential Evolution – Convergence Extension (DE-CE) is proposed as part of the anomaly-based IDS for dimensionality reduction and feature selection. The new variant of DE adopts a new mutation strategy that ensures the continuously generating new solutions for the current population, thus ensures selecting the most relevant features from the dataset. The well-known NSL-KDD dataset is adopted for training and testing the proposed anomaly-based IDS. Evaluation is performed against previously proposed DE algorithms with different mutation strategies and PSO in terms of number of selected features, Accuracy, False Positive rate (FPR), recall, and precision for five different classifiers. The proposed DE-CE outperformed the classical DE and PSO algorithms in all performance evaluation metrics, where it achieved the highest accuracy of 99.4744% and lowest FPR of 0.3198%
MODELING AND OUTDOOR CHARACTERIZATION OF TWO THIN-FILM PHOTOVOLTAIC MODULES: A CASE STUDY OF A-SI AND CIGS MODULES IN A SEMI-ARID CLIMATE ZONE
No full textReliable modeling of photovoltaic (PV) module performance under field conditions is important for optimizing system design. This study evaluates two promising thin-film technologies, an amorphous silicon (a-Si) 5W module and copper indium gallium diselenide (CIGS) 7W module, under Baghdad’s semi-arid climate over six months. Outdoor current-voltage and power-voltage curves were measured at irradiation levels of 500-1000 W/m2 and module temperatures of 25-50°C. A validated five-parameter model was used to simulate module performance. At 1000 W/m2 irradiation and 25°C, measurements showed maximum power outputs of 4.05W and 7.11W for the a-Si and CIGS modules respectively. The model predicted these values within 5% (3.9W) and 4% (7.37W). Model errors were below 10% for all test conditions. Specifically, at 500 W/m2 and 50°C, measured powers were 1.75W (a-Si) and 3.11W (CIGS), while modeled outputs were 1.8W and 3.07W respectively. This analysis provides a comprehensive long-term experimental dataset and accurate modeling of these thin-film technologies under real Middle Eastern operating conditions. It can aid optimization of PV systems based on these promising thin-film technologie
APPLICATION OF YOLO MODELS IN THE DETECTION OF FISH BEHAVIORAL CHANGE UNDER ACUTE EXPOSURE TO SYNTHETIC ESTROGEN IN THE ENVIRONMENT
No full textChanges in the behavior of small fish have recently been commonly used in assessing the impact of water pollutants, especially those of the group of endocrine disruptors. Behavioral studies mostly use visual observations, which can introduce bias and inconsistency in observational results. Recent studies have developed computer vision tools for tracking fish movements that allow automatic detection of small fish movements with high accuracy and consistency. In addition, computer vision combined with machine learning can help analyze, identify, and predict changes in fish behavior, easily integrated into environmental and ecological monitoring systems. This study uses YOLO (You Only Look One) algorithm models to detect fish in video data. Comparing the effectiveness of YOLO versions with the training data set shows that the YOLOv8s model has the highest efficiency and is selected for detecting and analyzing fish behavior in the environmental impact assessment model. The amount of image data for training the YOLOv8s model is also determined to be approximately 800 images. The training results show that YOLOv8s has high detection efficiency with a high frequency of detecting 11 fish in video frames. Results from detecting and analyzing fish positions in video data using the trained YOLOv8s model showed that the males of both mosquitofish (Gambusia affinis) and medaka (Oryzias latipes) species were affected following a two-day acute exposure to the estrogenic stressor, 17a-ethinylestradiol, in the aquatic environment at a concentration of 5 ng/L. While male mosquitofish when not exposed to estrogen tended to pay more attention towards the tank compartment containing female medaka, when exposed to estrogen, they increased their tendency towards the compartment containing male medaka. Additional research is needed to increase the accuracy and effectiveness of the YOLO algorithms in fish detection for behavior evaluation in an environmental impact assessment model
STUDIES OF MECHANICAL PROPERTIES OF BOTTOM ASH FILLED PBT COMPOSITES
No full textParticulate filled polymer composites have been very useful due to their low cost and wide engineering applications. Fly Ash as well as Bottom Ash emerging from thermal power plants have been an inevitable environmental hazard. Polybutylene Terephthalate (PBT) polymer has outstanding mechanical, electrical, chemical properties, better dimensional stability and low moisture absorption properties. Therefore, it has been very widely used for automobile and industrial applications. PBT composite of Bottom Ash, with and without a coupling agent has been made in this work. It is found that flexural modulus and flexural strength of this composite have increased for 90 microns Bottom Ash with the 15% (by wt%) loading and also a cost saving of 15% is reported. However tensile strength and impact strength are found to be decreasing marginally with the increase in % loading of Bottom Ash in the polymer. Scanning electron microscope (SEM) was used to characterize the samples for microstructure
A BIM-BASED SYSTEM FOR EARNED VALUE ANALYSIS IN CONSTRUCTION PROJECTS
No full textTraditional methods of construction monitoring rely heavily on manual data entry and 2D drawings, leading to significant inaccuracies and inefficiencies. These challenges include difficulties in tracking progress, potential human errors, and limited alignment with earned value analysis (EVA). Building Information Modeling (BIM) is expected to address these issues. This paper proposes a BIM-based Earned Value Analysis System (BEVAS) which comprises three modules: 1) BIM Model Requirement module specifies that the 3D model should be validated using clash detection and contain parameters for progress input; 2) Progress Data Entry module describes how progress data is inputted, while 3) Earned Value Analysis and Report module details the process of transferring from the 3D model to project control databases and analyzing it using earned value analysis, generating visual and quantitative reports. BEVAS is validated through a case study of a 3-story building, demonstrating its capability in both progress monitoring and earned value analysis. This system has shown significant improvements in monitoring accuracy and ease by providing a reliable earned value analysis. It contributes valuable insights for better decision-making in construction projects
INTELLIGENT PLASTIC BRAND AUDIT FOR EXTENDED PRODUCER RESPONSIBILITY INITIATIVES USING MACHINE LEARNING MODEL
No full textThe ever-growing volume of plastic waste poses a significant threat to global ecosystems. Existing waste management systems often struggle with the identification and sorting of plastic waste due to limitations in the scalability and cost-effectiveness of smart technologies. One key aspect of plastic waste mitigation is to enhance extended producer responsibility (EPR) capacity through plastic waste auditing. While computer vision techniques have been explored for general waste sorting, there is a lack of research focused on automated brand identification within plastic waste. This paper proposes a novel Intelligent Plastic Brand Audit (IPBA) system leveraging TinyML with machine learning capabilities for resource-constrained edge devices. The system utilizes a lightweight Faster Objects More Objects (FOMO) model trained on user-generated labelled photos and data. The performance evaluation of FOMO for IPBA was performed on two hardware platforms: Arduino Nano BLE and ESP32-EYE-CAM. Across both configurations and with five plastic brand classes, the system achieves high accuracy, with a minimum F1 score of 93.5%. These results indicate the potential of IPBA to improve existing manual sorting systems and support circular economy initiatives. By facilitating automated brand identification in plastic waste, IPBA can enhance EPR programs and hold brands accountable for their waste footprints