Honours and Masters project

Are you interested in applying your database knowledge to a real project? This is a collaboration with the Faculty of Medicine, Monash University.

Are you interested in working with hospital data? This project is a collaboration with the Faculty of Medicine, Monash University. In this project, you will be working with medical doctors from Monash Health. 

Large language models (LLMs) have recently made significant progress in machine translation quality [1], but they still struggle with maintaining consistency and accuracy across entire documents. Professional translators commonly use translation memory (TM) tools to reuse past translations, ensuring consistent terminology and phrasing throughout a document.

The research project aims to investigate:

- Multi-Model Fusion with Deep Neural Networks for Future Energy Systems (Smart Grid). 

Future energy systems are envisioned to be running decentrally with full automatic control, high proportion of renewable energy (e.g., wind & solar), and abundant storage facilities. With many types of renewable energy sources are weather and climate dependent, accurate and timely prediction on reliability risks (e.g., loss of generation, voltage issues, and thermal limit violations) due to weather/climate are often necessary.

The rapid growth of electric vehicles (EVs) is transforming the transportation systems worldwide. Both EV fleets and private EVs are emerging as a cleaner and more sustainable component of urban mobility, forming an effective way to solve environmental problems and reduce commute costs in future smart cities. Due to the complex spatiotemporal behaviors of passengers and their travel patterns, the unmanaged electric charging demand from EVs may significantly impact the existing transportation and electrical power infrastructure.

Thanks to the widespread deployment of smart meters, high volumes of residential load data have been collected and made available to both consumers and utility companies. Smart meter data open up tremendous opportunities, and various analytical techniques have been developed to analyse smart meter data using machine learning. This project will provide a new angle toward energy data analytics and aims to discover the consumption patterns, lifestyle, and behavioural changes of consumers.

Traditional active learning helps reduce labeling costs by selecting the most useful examples from a large pool of unlabeled data. However, in many real-world cases, such a large pool doesn't exist or is expensive to collect. This project explores a new approach using large language models to create synthetic unlabeled text data instead. Rather than just picking data to label, the model will also generate new examples that are diverse and potentially helpful for learning.

This project focuses specifically on LLM applications: chatbots used in customer support (e.g., healthcare). The goal is to investigate how user requirements (e.g., “the bot should de-escalate frustrated users”) can be systematically translated into prompt templates or prompt strategies.

This project aims to develop privacy-preserving deepfake detection techniques that enable accurate and secure identification of synthetic audio and video content without exposing sensitive user data. Traditional detection methods often require access to raw audio or visual inputs, raising significant privacy concerns, especially in scenarios involving personal or biometric data.

Mis/disinformation (also known as fake news), in the era of digital communication, poses a significant challenge to society, affecting public opinion, decision-making processes, and even democratic systems. We still know little about the features of this communication, the manipulation techniques employed, and the types of people who are more susceptible to believing this information.

This project extends upon Prof Whitty's work in this field to address one of the issues above.