Haifa’s project aims to enhance cyberbullying detection by exploring the causative factor of cyberbullying spreading, known as Bystander Contagion. This often-overlooked aspect plays a crucial role in the escalation of cyberbullying behavior. Haifa plans to investigate the detection of cyberbullying through the evaluation of users’ profiles, activities, and social media history, with a specific focus on bystander contagion. The goal is to identify influential cyberbullies or bystanders, enabling the suspension of their accounts to effectively reduce cyberbullying incidents.

In developing the cyberbullying detection model, Haifa will address key points, including the influence and characteristics of bystanders, measurement of bystanders’ contagiousness in cyberbullying diffusion, and identification of the most influential users capable of influencing others. The project will utilize datasets from the Twitter platform, with a focus on Arabic language tweets, acknowledging the unique challenges posed by the complex nature of cyberbullying in Arabic. Haifa’s methodology involves labeling and annotating the data, preprocessing procedures, and applying natural language processing, sentiment analysis, and machine learning with feature extraction for classification purposes.

Aisyahtul was a graduated Master student under the supervision of Dr. Syaheerah and is continuing with her PhD. Her Masters project focuses on data visualization embedded with time series prediction regarding the attrition and promotion issues among the permanent USM staff. The dashboard that describes the attrition pattern and promotion structure that happened in USM has been developed successfully using PowerBI. In order to touch the prediction element of data science, the attrition trend using the time series prediction models has been built using Python and embedded into PowerBI for the best visual. For her PhD, she is exploring multimodal bribery tendency prediction using eye-tracking and survey questions.

Samira Elsamad is now pursuing her PhD under the supervision of Dr. Syaheerah. Her work as a research assistant will focus on developing a machine learning model to analyze fMRI and EEG images to detect neural patterns linked to bribery tendencies. She will use deep learning techniques to train the model to automatically recognize distinct brain activity patterns when participants are exposed to bribery-related stimuli. By combining EEG’s fast electrical signals with the spatial detail from fMRI, she aims to create a predictive model that identifies neural markers associated with pro-bribery attitudes, providing insights into the neural basis of corrupt behaviors.

During her Masters’ degree, she was part of the data science team at the IMCC (International Mobility & Collaboration Centre) at USM working under Dr Syaheerah. Her work during her Masters focused on presenting an interactive dashboard that contains all relevant USM international network insights. Part of her work was evaluating the impact of USM’s partnership through new measures involving data scraping.
The goal of this project was to empower decision makers at the IMCC to formulate the most effective and timely internationalization strategies, and provide a concrete way to evaluate past strategies.

I am Assad H. Thary Al-Ghrairi, currently a PhD student in the School of Computer Sciences at USM. I am also a member of the teaching staff at the College of Science at Al-Nahrain University, Iraq, Baghdad. I am interested in research in computer science, pattern recognition, image processing, artificial intelligence, multimedia, remote sensing, and satellite image classification. Deepfake technology has advanced significantly, posing challenges in distinguishing real from manipulated content, particularly in facial images and videos. Artificial intelligence (AI) plays a crucial role in detecting deepfakes through techniques such as Generative Adversarial Networks (GANs) and Convolutional Neural Networks (CNNs), which analyze facial features and patterns for inconsistencies. Key indicators of deepfake content include unnatural facial movements, artifacts in skin tone, and audio-visual synchronization issues. However, the evolving nature of deepfake creation tools makes detection increasingly difficult, necessitating continuous adaptation of detection methods. Effective AI models require extensive datasets of both real and deepfake content to improve accuracy. As this technology progresses, ongoing research and development in detection strategies are essential to mitigate the risks associated with manipulated media.

Badr’s research interest lies in biometrics: particularly the study focuses on unconstrained facial recognition – occluded images with additional challenges such as expression, variations and single sample per subject. Badr has attempted to study occluded images coupled with expression, a new challenge few research in this area considers.

Hazem, an entrepreneur, MSc student at USM, and co-founder of YourVoice, enabling the blind to control their devices by voice.

He earned his BSc in Computer Science (Honours) in 2024, majoring in Intelligent Computing. During his journey, Hazem gained expertise in AI through internships, including ViTrox Corporation, where he applied state-of-the-art computer vision to detect semiconductor defects. His award-winning Final Year Project developed tools to filter inappropriate web content for children.

Currently, his research focuses on automating bystander stance labeling using Large Language Models (LLMs) to improve cyberbullying detection.

Alongside his research he has co-founded his startup YourVoice, in October 2024. Where it is a software that enables the blind and disabled to work, study and have fun online by naturally conversating with YourVoice and it executes then replies to them the results without needing to look at the screen. It comes with its own rehabilitation program to prepare the blind and disabled for the real-world.

MSc. Mix Mode

Michael Kong received his Bachelor of Computer Science with distinction from the University of Wollongong in 2016 and his Master of Science (Computer Science) from Universiti Sains Malaysia in 2018. His research for his masters thesis focused on image recognition in the area of microbiology, particularly for cyanobacteria recognition. In his free time, he enjoys learning about new and upcoming web technologies as well as photography.

Najlaa Sadiq Mokhtar is a MSc. candidate who embarks on an interesting research in the field of Human Computer Interactions, focusing in Affective Computing (personality, behaviour and emotion analysis) in learning, which. In short, she attempts to investigate the effects of emotions in learning. She hopes to build an affect-sensitive synthetic tutor that could detect and respond appropriately to learner’s state. The emotions involved would be those that accompany learning. Najlaa is now a senior software engineer in a private company.

Laura Ren’s interests lie within the area of Social Informatics, which is the study of social phenomena, behaviours and structure through the use of technology. Her work focuses in interaction in the Chinese cultural context, using social media features. Currently she is a MSc. Informatics candidate at PPSKOMP, USM. The idea of making machines understand people’s language and emotion has attracted her attention. She is eager to explore the new world, as Tagore said, she believes this would be the new power of her growth.

Muhamad Aizat has developed a machine learning project that will detect a pin failure from testing product data that focused on enhancing current processes which is manual effort of debugging why the pin was fail and replace the pin. The process is lengthy where during the events of products fail in testing, a technician who is debugging the TIU will first address the issue by manually locate the failure pins. The criteria of the pin failure will be determined by test program, and it’s embedded into software that the machine runs. While testing, if the scenario happens, the software will notify user to check the pin of TIU if there is any pin failure. After that the technician will debug the pin and replace it. To improve this process, the project embedded machine learning to analyze the data by predicting the failure before it actually fails. It will tells the user that certain criteria of data will likely fail the testing and notifies the location of pin that going to cause the failure.

IMCC Bantu aims to revolutionize the support system for full-time international students by leveraging machine learning techniques to match them with the perfect helper. Zita’s innovation addresses the critical need for personalized assistance during the challenging transition to a new educational environment.
Machine learning algorithms analyze various parameters, including cultural background, language proficiency, to identify the most suitable mentors or guides for incoming international students. This tailored matching process ensures that newcomers receive guidance and support from someone who understands their unique circumstances and can provide relevant assistance.
This initiative is crucial as it fosters a smoother integration process, easing the students’ adaptation to a new country. By utilizing machine learning, the IMCC Bantu project facilitates connections based on shared backgrounds that will help reducing the stress often associated with studying abroad. Ultimately, this personalized assistance contributes to their overall well-being and successful integration into their new academic community.

The aim of Belén’s FYP is to reduce cyberbullying on Twitter by taking into account bystander roles and the effect of their contagiousness.