Search

Sentiment analysis (SA) is the study of people's emotions and attitudes toward a particular topic. It is beneficial for monitoring and analyzing social media text in order to gather public opinion. Despite the fact that there are SA applications for monolingual text such as English and non-English languages like Hindi, Chinese and French, the Malay language has far fewer works, not to mention the mixed language such as Malay-English (also known as Manglish). Other than comments and posts from websites and social media, the emoji used by internet users can also help to provide better insights into how they truly feel about a particular topic. Our work focuses on Malay-English mixed language comments and posts on how Malaysians feel about daily new cases of Covid-19 in Malaysia. We proposed a neural network framework to perform SA on languages spoken by Malaysians, namely Malay, English, and Malay-English, by also taking into account the emoji used by internet users. The data was pre-processed to remove noises and then transformed into word vector representation using word embedding technique. Then we propose a framework that involves training and testing mixed language textual data along with emoji analysis by using bidirectional Long Short Term Memory (biLSTM) neural network. To compare with the proposed method, several machine learning models and Long Short Term Memory (LSTM) with word vectorization was used. Finally, compared to the machine learning model such as Naïve Bayes and Logistic Regression, neural networks such as LSTM, the proposed method; biLSTM with tuned hyper-parameter for Malay-English mixed language achieved the highest accuracy of 76.6%, and macro F1-score of 69.6%. [ABSTRACT FROM AUTHOR]

The paper presents the design and investigation of a wearable textile antenna (receiver) and transmitter antenna operating in the wireless body area network (WBAN) of 26 GHz band for 5G mobile networks. The wearable antenna with an overall size of 30 mm x 40 mm x 1.26 mm achieves good impedance matching, high gain, and directive radiation pattern. Both antennas were designed using CST Microwave Studio to validate the simulation results. A rectangular radiating patch comprises a Shieldit electrotextile situated on one side of a non-conductive substrate panel with the ground plane. The bed sheet cotton fabric is used as the non-conductive substrate due to its widespread use in daily clothing with a dielectric constant is 3.2 and the loss tangent is 0.0027. In addition, the wearable antenna successfully achieved the high gain and efficiency of 12 dB and 90.83% respectively. Moreover, the antenna operating at 26 GHz with -40.48 dB return loss, which is less than the -10 dB in requirement. The simulated results show that this proposed wearable antenna is best suited for wireless body area network applications. Hence, the wearable antenna is simple, compact and easy to fabricate. [ABSTRACT FROM AUTHOR]

Relatively more smartphones developed nowadays come with GPS location software. GPS information is fetched from the satellites and can be retrieved from Bluetooth, wi-fi, cell-id, and a combination of those tools. GPS-based mobile apps have the features of fetching the location of a device using geo-coordinates and some measurement which is an added value in developing various software. Due to the pandemic disease of Covid-19, the world is now aiming for global digitalization. Countries such as Singapore, Chine, and even Malaysia developed applications that can help the citizen in obeying the SOP and receiving current and reliable information regarding the Covid-19 issue. Thus, a real-time monitoring application for PUS s beneficial to break the chain and contain the disease. This research has been done to develop a localization application for Person Under Surveillance using GPS mobile apps. The application implemented the Firebase Authentication in the sign-up and sign-in process and is developed using Android Studio IDE 4.1 integrated with cloud-hosted database Firebase Realtime Database. Android version 6.0 is used which is API level 23. Application is developed to have the functionalities of user registration, fetching coordinates, setting a boundary, and in-app alert. The application was installed and tested on the Pixel 3 API 23 emulator. [ABSTRACT FROM AUTHOR]

RFID technology is evolving as one of the most popular technologies in this era of technology, fast gaining attention due to high demand from users and rapidly garnering interest in scientific and commercial areas. The frequency used will be determined by the RFID application, and the power rate will change as the frequency increases. Without a straight line of sight, the RFID tag antenna can identify distinct objects. For wearable applications, several RFID tag antennas are too big to match the chip impedance. In order to overcome that problem, a small tag RFID tag antenna for UHF is designed which aimed to be operated from 865 to 867 MHz for assembling production. Impedance matching is used to transforms the impedance of the radiating antenna, to match the chip impedance. To design and simulate the designed antenna, CST Microwave Studio software has been used in this project to get the desired result which is the return loss and gain. The design for this antenna is very simple to ease the fabrication process. Overall, the construction comprises a spiral-shaped loop RFID tag antenna printed on Roger substrate RO4350 with a height of 0.8 mm and a dielectric constant of 3.3. The simulation result of the reflection coefficient of the antenna is 866 MHz at the operating frequency. [ABSTRACT FROM AUTHOR]

RFID is not a new technology. It has been applied in various industries such as for wearable applications. Common RFID tags especially for those that have been designed and are available are not independent of the incident receiver angle. Numerous wearable antennas on the market are only designed for a certain received angle. For example, a wearable RFID antenna is used in medical as a pulse reading detector. If the patient makes any movement, the patient's pulse reading is no longer accurate or there may be no pulse reading. Hence, the purpose of this project is to design and RFID antennas using Frequency Selective Surface, FSS for wearable applications that are independent towards the incident angle and small in size. In this project, several antennas design with Frequency Selective Surface (FSS) is proposed. The design for this antenna is round, square, and hexagonal. This antenna has an operating frequency from 2.4 GHz to 5.8GHz, bandwidth efficiency > 50%, dielectric constant 1.30, independent incident angle up to 60 degrees, and has a high gain of around 2 to 3dB. [ABSTRACT FROM AUTHOR]

The objective of this study is to implement machine learning (ML) to identify and classify the level of contamination in leftover cooked foods based on its aroma. An evaluation on the smell profiles using as a model local Malaysian lunch or evening foods that have always been stored as leftover cooked food is done in this study. To capture the data, a simple e-nose application is built and affixed to the food containers, which will accommodate four types of sensors sensitive to different gases and is programmed using the Arduino platform. To determine the aroma categorization of leftover Malaysian cuisine, samples are examined using RStudio. The results in this study demonstrated satisfactory performances by k-Nearest Neighbours (k-NN), Support Vector Machines (SVM), and Random Forest (RF) with accuracies ranging from 87.5% to 100% using the oversampling and undersampling techniques. Unfortunately, Linear Discriminant Analysis (LDA) gave poor performances (19.64% – 58.93%) in classifying the contamination level of the samples. Hence, the results obtained gave an indication that the electronic nose presented in this research was a promising for classification of contamination level for leftover cooked foods, allowing food to be better anticipated as to whether it is still edible or not. [ABSTRACT FROM AUTHOR]

The paper introduces a novel watermarking scheme to enhance robustness and integrity in medical images using the Knight's Tour algorithm. Medical image digitization and sharing necessitate robust techniques for data security. Existing watermarking methods often lack resilience against attacks and maintain imperceptibility. To address these issues, the proposed scheme employs the Knight's Tour algorithm to determine the optimal pixel locations for watermark embedding. Extensive evaluations demonstrate its effectiveness against geometric attacks, specifically image rotation and scaling. The study's significance lies in providing a reliable solution for safeguarding medical image data, ensuring authenticity, and bolstering trust in healthcare applications. Moreover, the scheme's adaptability extends its significance to multimedia data protection and content authentication. [ABSTRACT FROM AUTHOR]

The integration of Augmented Reality (AR) technology in indoor navigation has transformed spatial awareness and user interaction. This research focuses on "AR@Campus," an AR app designed for campus navigation, aligned with Industry 4.0 principles. It leverages AR to enhance human perception and dynamic digital interaction. The primary goal is to provide an intuitive campus navigation experience using IndoorAtlas positioning technology. Key features include overlaying virtual elements like paths and landmarks onto real-world views, improving spatial understanding. The study covers software development, interface design, and user evaluation, utilizing the IndoorAtlas SDK and Android Studio for precise positioning. The user interface prioritizes ease of use and personalized experiences. User testing is integral, confirming the app's effectiveness, praised for its user-friendly interface and spatial awareness enhancements. AR@Campus represents a significant step forward in campus navigation, benefiting users by seamlessly integrating AR technology into their daily experiences. [ABSTRACT FROM AUTHOR]

Plant disease classification is essential to the agriculture industry. In this work, a tomato disease classification using Extreme Learning Machine (ELM) with image-based features. Extreme Learning Machine (ELM), a classification algorithm with a single layer feed-forward neural network where it can be combined with few activation functions. The image features as the input where the image is pre-processed via HSV colour space and extracted using Haralick textures, colour moments and HSV histogram. The features are then loaded in the ELM classifier to perform the ELM training and testing. The accuracy result of ELM classification is then analysed after the validation. The dataset used for disease detection is tomato plant leaves which is a subset of the Plant-Village dataset. The results produced from the ELM demonstrate an accuracy of around 84.94% which is comparable to classifiers such as the Support Vector Machine and Decision Tree. [ABSTRACT FROM AUTHOR]