Dr Jasy Liew: Computers do help make the world a better place


  • People
  • Monday, 06 Nov 2017

Dr Jasy Liew Suet Yan wants to develop a computer system or model that can detect early signs of depression in a person. Photos: L'Oreal Malaysia

When Dr Jasy Liew Suet Yan was a teenager she became fascinated with the idea of working with computers.

“My older sister was studying Computer Science when I was in secondary school and during her breaks she would come home and show me the programmes that she had worked on. I thought it was so cool that you could create a system that would allow computers to do the things you want it to do,” said Liew, 32, in an interview recently.

Today, Liew is a senior lecturer at the School Of Computer Sciences, Universiti Sains Malaysia. She is also one of the three recipients of the 2017 L’Oreal-Unesco Awards: National Fellowship. The others are Dr Teh Su Yean and Dr Ho Weng Kee.

“I was shocked that I got it. I create computational models so I didn’t think I could win. I wasn’t sure if the jury would consider what I do as a science. I think there’s still a perception that science is limited to chemistry, biology and physics,” she said.

Fortunately, her proposal got the attention of the jury.

What is your research about?

Basically, I want to develop a computer system or model that could detect early signs of depression in a person, through the language they use in their social media postings.

Can we build a computational model that could automatically detect expressions of emotions among a chunk of text?

And since depression is associated with certain emotional patterns, can we then train computers to detect these patterns? Can we trigger alerts which lets a person know that they are showing signs of depression?

Why did you decide to do this research and topic?

For my PhD dissertation, I wanted to see whether we could train computers to tell us a wide range of emotions. So, I created 28 emotion categories for the computer to process, from sympathy and gratitude to hope, fear, happiness, sadness, anger and more.

This opened up a greater potential in the kind of application that we can do to figure out what people are feeling.

ho weng kee
Recipients of the 2017 L'Oreal-Unesco Awards: National Fellowship: (from left) Dr Jasy Liew Suet Yan, Dr Teh Su Yean and Dr Ho Weng Kee.

The depression part comes with a very personal story. When I first started my PhD programme in the United States, I was placed in a team of four people. We were asked to tell a professor what we were planning for our dissertation. He told us bluntly that some of our ideas were not good and that we had to work harder.

I think all of us got really harsh comments from him but we mostly took that as a lesson and moved on. Unfortunately, one girl took it really hard; she never got over it. I think it made her lose confidence and question why she was in that programme.

Soon, she started disappearing and none of us knew what was going on with her. Six months later, she fainted at school and was diagnosed with severe depression.

That was the first time I realised just how serious depression was. It was then that I began to think about using the emotion detectors in a way that could help early detection of depression.

What would be the biggest challenge in getting this research done?

Research work is never a smooth one-way trip! The biggest challenge would be ... when things don’t work out the way we thought it would. That means we need to explore other ways to make things work, maybe even start from scratch.

This research is also set up as a long-term project so funding is important. I am very grateful for the grant from the L’Oreal-Unesco For Women In Science programme because that helped us kickstart things. But we will definitely need more funding in the long run.

I have actually just started the project and have two final year project students working with me on developing the system. I recently brought in a psychiatrist to help us get access to patients who have been diagnosed with depression and also to provide us with other key information. He would be the domain expert.

What would you say to a young girl who has dreams of pursuing work in Computer Science just like yourself?

First of all, you can apply Computer Science knowledge into so many other fields of study, so it’s very interesting.

Computer Science also requires a lot of problem-solving skills. That’s something that anyone can do, it doesn’t matter what gender you are. I know there is a tendency for some people to say that science is too difficult for a girl to understand but that is wrong.

Don’t be discouraged by negative words, if you put your heart into it and work hard, you can achieve so much.


L’Oreal-Unesco Awards: National Fellowship

The fellowship is open to all Malaysian women researchers and scientists who are under the age of 40 years old. Applicants must either be PhD holders or are currently pursuing research studies in any scientific field.

Each recipient will receive a RM30,000 grant to help pursue their research.

Two jury panels were set up to comb through all the applications. The preliminary panelists for 2017 comprised eight scientists, all of whom are former grant recipients.

The main jury panel for this year was headed by Malaysian astrophysicist Prof Emeritus Datuk Dr Mazlan Othman, who is also the director of the International Council Of Science Regional Office for Asia and the Pacific. Other panelists are Associate Prof Dr Ramzah Dambul, the deputy secretary general, Ministry Of Science, Technology & Innovation (Science Division); Prof Dr Adeeba Kamarulzaman, dean of the faculty of Medicine, Universiti Malaya; and Prof Dr Norzulaani Khalid, director of UM Cares, Universiti Malaya.

Application to the programme usually opens in March every year. You can follow the L’Oreal-Unesco For Women In Science Facebook page or the L’Oreal Malaysia website for updates.


ALSO READ:

Dr Teh Su Yean: Using Mathematics to address an environmental problem

Dr Ho Weng Kee: Developing a breast cancer assessment tool


Article type: metered
User Type: anonymous web
User Status:
Campaign ID: 7
Cxense type: free
User access status: 3
   

Across The Star Online


Air Pollutant Index

Highest API Readings

    Select State and Location to view the latest API reading

    Source: Department of Environment, Malaysia