Thesis Title
Forecasting of Listener Reaction Trajectory in Dialogues

Research Description
Anticipating a listener’s emotional and behavioral reactions is essential for planning effective and empathetic speech. Existing studies overlook its unique properties, including strong dependence on individual listener traits, along with delayed and evolving responses. This project formulates the task of listener reaction trajectory forecasting, develops methods for personalized prediction, and explores its downstream applications such as improving emotional support conversations, predicting conversational derailments, and strengthening large language model safety. By enabling proactive anticipation of reactions, the proposed research aims to enhance the safety, effectiveness, and appropriateness of conversational systems.

Supervisors
Prof Chris Leckie, University of Melbourne

Thesis Title
Optimizing Information Retrieval in LLM-based conversational search engines by assessing physiological signals

Research Description
State of knowledge plays a vital role in determining what information is needed for a search task, and people learn from the information during the process to its accomplishment. This dynamic process can lead to diversity in queries and search results. With the increasing usage of Large Language Model (LLM)-based conversational search engines, it is crucial to regenerate new answers and suggest query reformulation or refinement options for effective Information Retrieval (IR). User feedback is invaluable for enhancing the IR process. This study aims to examine human searching behavior on LLM-based search engines. It aims to measure user feedback by analyzing physiological signals caused by a change in prior knowledge. The goal is to gain insight into people’s information searching (IS) behavior and seeks to evaluate the optimal timing for recommendations from the search engine.

Supervisors
Dr Damiano Spina, RMIT University
Dr Johanna Trippas, RMIT University
Dr Dana Mckay, RMIT University