Shining in an unusual time
Meet Gus, Lingyu, Lexie, Sophia and Hai Ho, Master of Data Science and Information Technology students who joined the MDAP team in mid-2020, shortly after we all started to settle into remote working life. Determined for distance to be no barrier, this intern group proceeded to contribute greatly to many of our collaborations. Here they reflect.
Gus (Hongwei) Yin
During my internship, I worked with the Melbourne Bioinformatics Training Group to develop a resource management system for internal and online resource sharing.
“I had no prior knowledge in web development, but with the help of my supervisors, I researched the best tools for this project: Django, PostgreSQL, HTML and CSS.”
This internship provided me with real-life experience in data science and highlighted to me that a data scientist should be capable of using all kinds of technologies to meet the needs of projects as the problem solve. It was great to discuss researchers’ needs with them and the experience helped me improve foundational communication skills for data scientists.
Throughout the internship, I was able to keep pushing myself to overcome several challenges and learn completely new knowledge. And, engaging with supervisors and external collaborators, I gained a deeper understanding of the industry. I will continue to expand my knowledge in the related fields and new technologies to become a better data scientist.
I helped the Melbourne Pollen team develop their new pollen count and forecast app, which helps people to better manage allergies across Australia. My main challenge was to rapidly learn and use the programming language and the app development technologies.
“To overcome this challenge, I learned new technologies by taking online courses and practicing my new skills on the app. For each component of the app, I tried different approaches to achieve the outlook and functionality and choose the best one as the final result.”
Aside from providing the pollen count and forecast, the app has other functionality such as displaying weather forecasts and provides a survey form, which can help people record their allergy symptoms. So, there is a lot to consider during the process of developing the app, such as how to make a user-friendly interface, and how to manage the private data, and how to ensure the user data is securely transmitted.
This internship is the most valuable experience I have had. I learned how to use new technologies and gained exposure to the whole software development process. Working in a small team helped to improve both my communication and teamwork skills. After finishing my internship, I will continue studying and practicing my skills.
Lexie (Yiran) Yao
I collaborated with Dr Martin Hirche, Department of Management and Marketing, Faculty of Business and Economics on a project that looks into the relationship between spatial distribution and market share for consumer packaged goods in the United States market. My task was to help develop a program that would automate and streamline the analysis and calculation of the data for the study.
I approached this task by breaking down the existing sequential program into reusable, modular functions. By adopting third-party packages that are designed to support the processing of large amounts of data, I was able to significantly reduce the data processing time. Next, I produced a detailed user guide to help the end-user understand the usage and logic of each function.
As today’s research continues to become more and more data-intensive, data analytics further becomes the cornerstone of an efficient and smooth-running research process. This experience has shown me that advanced technologies and robust analytics skills are keys to achieve that.
“It certainly was gratifying to witness first-hand how data science contributes to the advancement of various research domains. The MDAP team was exceptionally welcoming, friendly, and helpful, and each fellow intern I came to know was motivated and goal-oriented. I really enjoyed my time at MDAP.”
The past few months have been fruitful in that I have learned many new technologies, and I am interested to learn more about in-depth. I might invest some time in personal side projects that utilise those technologies.
During my internship, I mainly contributed to developing the Melbourne Research Cloud (MRC), which offers free on-demand computing resources to researchers at the University of Melbourne (and affiliated institutions). I also briefly helped to develop Unicode, a web app that aims to connect the University research and teaching community with internal coding expertise.
In the MRC project, I was tasked with deciphering various use-cases and scenarios that the researchers face while using the Melbourne Research Cloud to accomplish various research tasks. In order to identify the various use-cases, I put myself in the place of the researcher and then performed some tasks on the Melbourne Research Cloud, to get a better feel of the platform itself as this was my first experience working with a research cloud as well.
“I designed a survey for diverse researchers to participate in to better understand their needs and experiences. Over time, I identified some gaps and scenarios that will enhance the user experience of the researcher.”
The main insight I gathered from this internship is a better understanding of the entire data analysis process, from defining the problem to gathering data and finally analysing it to bring about new discoveries that help to address the problem. I had the opportunity to work with various tools that I had no prior experience of working with, such as Django, Anaconda and Qualtrics. I learned to accept new ideas and work diligently towards fulfilling them from the MDAP team and my fellow interns.
Next, I have some interesting project ideas to implement by using the various techniques and procedures that I discovered and learned during the course of this internship.
Hai Ho Dac
In my internship, I worked on the Video Wellbeing Assessment project, with the Melbourne Graduate School of Education (Centre for Positive Psychology). This project aims to digitise wellbeing and mental health assessment via short self-recorded video clips of people answering general non-sensitive questions and picking up emotion changes over time. In this project, we used computer vision technology to detect and track the changes of emotion, then export a summary analysis for researchers to evaluate the impact of school and workplace wellbeing programs.
This problem required a comprehensive knowledge of both wellbeing science and computer science. To meet the best of both worlds, I first researched relevant topics and previous works presented in academic papers and domain specialised articles. Once I got all the information and resources ready, I started building my first facial expression detector, capable of capturing a smile on people’s faces. This facial expression detector set a baseline for me to layer over (and compare) sophisticated models and then an emotion detector.
Importantly, we needed to detect negative emotions as it reflects a trajectory of depression. To do this, we focussed on identifying (and differentiating) three out of seven detectable emotions: happy, neutral, and sad.
“The result of our project demonstrates how artificial intelligence can help to automate the diagnosis of wellbeing and mental health assessment.
It also shows a potential for further integration of computer vision and other technologies, such as voice recognition (track the changing of the tone) or word analysis, to help improve the evaluation confidence of the application.”
During my internship, I have been given many opportunities to expand and push my ability to the limit, via many challenges in the project. Also, engaging with the external researchers allowed me to learn more about their domain knowledge and understand more about the research questions we were trying to solve.
I have developed my confidence, knowledge and skills. Thank you, MDAP!
MDAP’s internship program runs each quarter, all year round. If you are interested in learning more about it or would like to be notified about upcoming opportunities, please email us or keep an eye out on Twitter.