Bio

I am a postdoc research fellow @ CSIRO’s Data61, the national scientific research institute of Australia. I am working with Dr. Cecile Paris, Dr. Sarvnaz Karimi, and Dr. Diego Molla-Aliod in the Language Technology Team. I acquired PhD 🎓 from the University of Melbourne, supervised by A/Prof. Nicholas Geard, Prof. Karin Verspoor, Prof. Justin Zobel, and Dr. Benjamin Goudey.

My research focuses on the practical applications of NLP and large language models (LLMs) to address real-world challenges. I’ve developed automatic solutions for extracting key information from electronic breast cancer follow-up treatment records 📄, accelerating clinical research and decision-making. I also created an NLP-based automated quality assessment system for biological databases 🧬, enhancing the reliability of gene and protein function knowledge to support protein research. Additionally, I proposed an innovative approach to integrate bio-ontologies into machine learning frameworks, significantly improving the accuracy of gene function data quality assessments 🔍.

At CSIRO, I led efforts to improve the explainability of LLMs safeguard🛡️for suicide ideation prediction on social media. I also lead a team to work on the comprehensive evaluations of LLMs on a range of affective computing tasks, including cross-lingual and cross-cultural emotion detection and temporal sentiment regression. I provide supervision to researchers from corporate/major government of other countries, supporting projects that involve close cross-national, cross-institutional, and cross-domain collaboration. Currently, I’m working on a scientific exploratory research project, aiming to assisting clinical psychologists in the discovery of novel emotion regulation patterns from large-scale social media data and compose effective strategies to help people build resilience for mental health ❤️. We live in an era which encourages people to use emojis.

📩 Contact Me in English or 简中

Experiences 🤓

Service and Leadership ☸️

Miscellaneous

Where to CatchMe ☕

upcoming

past

Visitors of this site

Flag Counter Flag Counter