PhD Student SU Dan Leads HKUST Engineering Team in Award-Winning Research Related to COVID-19
SU Dan, a PhD student of Electronic and Computer Engineering (ECE), has recently led a team at the Hong Kong University of Science and Technology (HKUST) to win Kaggle’s COVID-19 Open Research Dataset Challenge with a question-answering system for medical professionals to quickly find answers to their queries about the pandemic. As early as her first month as a PhD student last year, she led another team to garner the Chatbot Millionaire Challenge organized by the Hong Kong Science and Technology Parks Corporation. The high achiever, who does well with her studies and in the workplace, is so down-to-earth that she talks about her phenomenal achievements as if they are commonplace.
Extraordinary abilities at a young age
Unlike many high achievers who consciously set far-reaching goals to help them overcome hurdles on the way, Dan is unique in the way that she does not seem to need clearly-defined goals to become motivated. The naturally driven, intelligent and diligent student from Shaanxi in China started school one year earlier than other kids of the same age. This did not stop her from always coming first in class. She won awards in national competitions in mathematics multiple times, and was able to get good score among junior high students when she was only a pupil in primary five.
Her world of knowledge was further widened as she went to an elite senior high school in the same province. There, she had the opportunity to join other outstanding students and meet professors from the prestigious Tsinghua University.
Due to little prior exposure to the computer however, she almost had to start from scratch when she joined the University of Science and Technology of China’s Computer Engineering program. While this made her rather different from other classmates, it did not take long for the talented and hardworking student to be awarded scholarships again and demonstrate leadership abilities in the university’s career development society.
After some struggles during the fourth year, the top student declined an offer at the National Engineering Laboratory for Speech and Language Information Processing (NEL-SLIP), and instead applied for universities in Hong Kong and Singapore. Joining MPhil in ECE at HKUST in 2009 with scholarship, she published three papers on personalized music emotion classification via active learning and participated in conferences abroad, all of which were notable for an MPhil student. And despite being busy with her studies, she managed to fulfill responsibilities as head of HKUST’s Mainland Students and Scholars Society to organize activities for fellow students.
Motivation from within
“I seldom thought about how I got to where I am today, until being asked during this interview. Somehow I didn’t really set long-term goals. I always try to do things well and solve problems on the way; it is natural. By doing these, opportunities came, and I grasped them,” noted the humble, high-achieving student.
“There might be painful experiences that drove me to tears, but the next morning I forgot about them and started anew. I am optimistic; hurdles are only meant to be inspirational. Always recalling the positive experiences, I feel grateful towards people who have helped me.”
Dan’s MPhil and PhD supervisor Prof. Pascale FUNG remarked, “I could see the potential in her, and I still think that she has more to offer now that she is already on the way to become a strong leader. As a bright problem-solver, she can see the bigger picture.”
Launching an exciting career
Upon completion of MPhil in 2012, Dan decided not to pursue PhD immediately. She launched her career by joining a well-established bank, but unfortunately the job did not offer enough challenges. In 2014, she co-founded a robotics company in Shenzhen.
The venture that followed involved SF Technology, an affiliate of the delivery giant SF Express. As the Senior Machine Learning and AI Engineer, she was given the arduous task to build a system with the aim to retrieve lost claims. “As an issue beyond mathematics, it could not be solved by artificial intelligence alone. Since it had to do with overall logistics and management, a credit system needs to be set up,” the then new employee quickly concluded.
Whereas other new recruits might have found it a time-consuming task with little promise of being successful, Dan once again showed her first-rate intellectual ability and perseverance by conducting substantial research on her own to build machine learning algorithms. “In machine learning, you have to get your hands dirty with data,” she stressed.
As a result, her boss was so impressed by her conceptual model that she was invited to meet with senior management and present her ideas. Soon after, she became the project lead and was given seven team members under her to achieve the company’s goals. Adaptations of the system, which helped to enhance risk control, are still being used today in one form or another. As part of the celebrated success, she acquired a few patents for the invention.
Repeatedly being given A+ for her performance during annual evaluations and being awarded the High Potential and Performance Award, she had set such a superb role model that she represented the company to go to MIT, Georgia Tech, Boston University, HKUST and other prestigious universities to recruit new talents.
At the peak of her career in the multinational corporation in Shenzhen, Dan made the life-changing move to return to HKUST in Spring 2019. “There is always the psychological complex in me to pursue PhD studies. If I let it go, I may never have the chance to pursue it.” So here she comes, pursuing PhD studies and leading HKUST teams to achieve victories at competitions.
Once again, Dan is ready to overcome challenges in relation to her uncommon choice. “Whereas my parents back in Shaanxi were worried that I might not be able to pay off my mortgage for the purchased flat in Shenzhen, I believe there must be a way and a solution. I might be slightly older than other PhD students, but I have learned to solve problems. Other people may feel the pressure, but I like finding solutions.”
Two weeks after her return to her alma mater last year, she was encouraged by Prof. Fung to lead a team, made up of PhD students and engineers from EMOS Technologies, to participate in the Chatbot Millionaire Challenge organized by the Hong Kong Science and Technology Parks Corporation. With merely a fortnight to prepare, and with all team members being simultaneously occupied with other projects, Dan led the team to quickly explore the rules and formulate a strategy. The team, working mainly on weekends and sometimes meeting online, ended up winning by a huge margin among competing teams specialized in data science and chatbot solutions.
Most recently, she led the team at HKUST’s Center for Artificial Intelligence Research (CAiRE) to achieve the highest scores, among over 1,000 global teams, regarding one of the 10 tasks of the COVID-19 Open Research Dataset Challenge (CORD-19 Challenge). The team built their own CAiRE-COVID system, a machine learning-based system with top-notch natural language processing (NLP) question-answering (QA) techniques combined with summarization for mining available scientific literature. Quickly generating ranked lists and paragraph-level summaries among over 57,000 scholarly articles about COVID-19 and related coronaviruses, it is meant to facilitate the medical community to find answers to queries and ultimately a cure. The AI Challenge is jointly organized by Allen Institute for AI, Chan Zuckerberg Initiative, Microsoft Research, Georgetown University’s Center for Security and Emerging Technology, the US National Institutes of Health’s National Library of Medicine, and the White House.
Building her portfolio in a male-dominated profession
As a female engineer and a PhD student, Dan has to consistently overcome hindrances regarding expectations from the family and the society. Having higher aspirations towards studies and career than her parents could ever imagine, she always had to choose the road less traveled and pursue her own way. “Individual thoughts were something I have had since a young age. I am grateful for the encouragement of Prof. Fung who has influenced me immensely,” she said.
Prof. Fung commented, “Girls are often discouraged to pursue PhD. For Dan, her strong motivation is all internal. Her precious strengths are not only about intellectual and technical skills; her leadership abilities are unusual. She had led two teams to win two competitions within only weeks of preparation.”
“Don’t think too much. Just do it.”
Interestingly, the down-to-earth talent does not have lofty goals. Instead, she takes things one at a time with a utilitarian, step-by-step approach. “Don’t think too much. When there are good projects, just do it and solve problems along the way. Be courageous.”
She is not overly pragmatic either. “Some people ask me how many awards I expect to get, but I never see things that way. It is important to filter out these noises. Just focus, try your best, work seriously and it will be fine. Never give up; you only live once. And I don’t like to admit failure.”
- SENG news (Jul 3, 2020): Prof. Pascale FUNG Helps to Debunk COVID-19 Myths with Science
- ECE news (Mar 7, 2020): A Question Answering and Summarization System Developed by CAiRE Team Won one of the 10 Tasks of The COVID-19 Open Research Challenge (CORD-19)
- ECE news (Mar 1, 2019): EMOS Technologies Won The Chatbot Millionaire Challenge in "AI+U: Explore and Experience Exhibition" Organized by HKSTP