Yueqing Xuan

About Me

I am a PhD student from the School of Computing Technologies at RMIT University and student member of Australian Research Council (ARC) Centre of Excellence for Automated Decision-Making and Society (ADM+S). I started my PhD in August 2022 and expect to complete it on or before July 2026. I work alongside my supervisors Assoc Prof Jeffrey Chan and Prof Mark Sanderson on explainability and fairness in Machine Learning and Recommender Systems. Specifically, my research aims at developing a fairness-aware and privacy-preserving recommender system, incorporating the use of machine learning and natural language processing (NLP) techniques. I am currently working on a project with a focus on analyzing user bias and item bias in recommender systems.

I have recently completed a master's degree in Information Technology at the University of Melbourne, with specialisation in Artificial Intelligence (AI). I have completed a research project and finished my master's thesis that is concerned with NLP. Specifically, we proposed a multi-modal Entity Linker that automatically recognises the object(s) of interest in the natural language commands. It is also capable of linking the textual mentions of entities to visual mentions of entities that correspond to the same real-world entities. I received a grade of 91 (First Class Honours / H1) for the research project. More details about my work are described here.

Upon finishing my research project, I continued working on the (parent) research project Humanoid robotic arm control through verbal commands as a casual research assistant at UniMelb. This project intends to develop a framework allowing people living with upper-limb disabilities to manipulate a robotic arm to perform various reaching and grasping movements. My work included deploying an off-the-shelf voice activation engine and an Automatic Speech Recognition (ASR) system for the robotic arm framework. I have further integrated the framework with a humanoid robot - Reachy robot, for real-time control using voice commands. I was also responsible for controlling and interacting with Reachy to carry out various verbal commands in the lab environment.

Education

• Supervised by A/Prof Jeffrey Chan and Prof Mark Sanderson.
• Scholarship recipient of the ARC Centre for Automated Decision-Making and Society (ADM+S).
• Research focuses on explainability and fairness in Machine Learning and Recommender Systems.

Specialised in Artificial Intelligence

• WAM: 84.50
• Relevant Coursework: Research Project, Natural Language Processing, Computer Vision, Statistical Machine Learning, etc.
• Achievements: Recognised on the Dean's Honours List 2021 at the University of Melbourne for top 5% students. Awarded Melbourne Graduate Scholarship 2022.

Major in Accounting and Information Systems

• WAM: 74.50
• Relevant Coursework: Database Management and Big Data, Information System in Business, Business System Analysis, Programming for Business, etc.
• Achievements: High Distinction (HD) average in programming subjects, Distinction (D) average in accounting subjects, Second place in COMP1531 (Software Engineering Fundamentals).

Publications

Peer Reviewed Papers

• Yueqing Xuan*, Edward Small*, Danula Hettiachchi* and Kacper Sokol*. "Helpful, Misleading or Confusing: How Humans Perceive Fundamental Building Blocks of Artificial Intelligence Explanations". In: Human-Centered Explainable AI (HCXAI) Workshop at CHI. 2023. [Slides]

Professional Service

Conference Reviewing

• SIGIR (2023).

Academic Projects

• Proposed a multi-modal entity linker that recognises the target and/or receptacle objects within verbal commands through machine learning, and performs visual-textual entity linking. [GitHub link]
• Performed end-to-end evaluations on the voice-controlled robotic arm control framework.
• Supervised by Dr. Lea Frermann and Dr. Adi Tarigoppula.
• Achieved 91 (First Class Honours) in this research project.

• Built a rumour classification model via BERT in PyTorch for detection of rumour tweets.
• Applied the classifier to analyse the characteristics of COVID-related rumours on Twitter.
• Ranked in the top 15% on the leaderboard.

• Implemented an autonomous agent based on reinforcement learning, which can play in the game Sequence.
• Ranked in the top 10% in the final competition.

• Created an agent-based model in NetLogo to investigate the role of asymptomatic individuals in spreading COVID-19.
• Designed experiments to explore the effectiveness of isolation for disease control.
• Achieved full marks in this project.

• Developed an Event Management System in Python and HTML that features event creation and registration for UNSW staff and students.
• Achieved High Distinction in this project.

Work Experience

The University of Melbourne, Melbourne

• Worked on the research project: Humanoid robotic arm control through verbal commands.
• Successfully integrated the Porcupine wake word engine and IBM Watson speech recognition system into the robotic arm control framework, and integrated the framework with the anthropomorphic robotic limb.
• Controlled and interacted with the humanoid robot for real-time control using verbal (natural language) commands.
• Demos of the experiments conducted in the lab environment can be found here.

The Altius Group, Sydney

• Effectively managed, configured, and troubleshoot a range of IT-related systems.
• Responsibly implemented and administrated database documentation, guidelines, policies, and procedures.
• Performed application and hardware configuration deployment and troubleshooting.

Shanghai Institute of Technical Physics of the Chinese Academy of Sciences, Shanghai

• Worked effectively in the Big Data team to analyse Geographic Information System (GIS) data by utilising the ContextCapture software.
• Designed and maintained database architecture, data structures, tables, dictionaries, and naming conventions to ensure the accuracy and completeness of all data master files.
• Performed the operational establishment and preventive maintenance of backups, recovery procedures, and enforced security and integrity controls.

Skills

Technical Skills

• Python libraries for machine learning: Frequently use PyTorch, Keras, OpenCV, Scikit-learn, NumPy, and Pandas.
• Programming Languages: Proficient in Python. Confident in JavaScript and Java. Familiar with C, SQL, and NetLogo.
• Business Intelligence Software: Familiar with SAP Lumir, SAS, and Power BI.

Language

• English: Pearson Test of English (Academic) - achieved 90 overall (IELTS Equivalent: Band 9), at least 83 for each of the 4 test components.
• Mandarin: Native speaker.

Personal

• Strong teamwork skills: Working smoothly and effectively as a team member through collaboration, effective communication, and providing assistance in a positive manner.
• Problem-solving skills: Developing both creative and practical solutions to solve identified issues and discrepancies, and contributing to effective outcomes.
• Time management skills: Managing time effectively through prioritising and making judgments about what is urgent or important, and coordinating different tasks to meet deadlines.