Elaman Earns His PhD

On behalf of the community, we congratulate our resident and dsml reading club research seminar organizer, Elaman Abdullin on successfully defending his PhD at Macquarie University in Sydney! Elaman defended a thesis titled "Eliciting Information From Users in Goal-Oriented Dialogue Systems" and received an award without corrections immediately after the defense (whereas even after a successful defense, candidates are often asked to make some changes to their dissertation).

Elaman studied synthetic dialogue generation, training models to ask questions using reinforcement learning, and the use of knowledge graphs. His work focuses on how LLMs can ask users questions to gather information for solving a task. He explored this through optimization modeling: a person describes their situation in everyday language, and an AI assistant helps clarify their goals, constraints, and available resources. We’re very happy for Elaman and wish him continued success!