Two students in the McGill Desautels Master of Management in Analytics (MMA) program have developed a to track COVID-19 cases in Canada.
In early March, Venkatesh Chandra and Ramy Hammam were closely monitoring the evolution of COVID-19 cases in Canada. With an interest in data analytics, the duo began building basic charts to track cases and find trends across different provinces.
Chandra and Hammam were surprised that there were no websites that provided a big picture view of how the situation was unfolding in Canada, so they set out to create a simple, easily accessible dashboard that reported on different metrics.
“As the numbers were slowly picking up and circumstances were deteriorating in Canada, we wanted to build a dashboard that would provide a snapshot of the situation,” says Chandra.
Within 24 hours, they had created a dashboard to monitor and visualize COVID-19 cases across Canada. Both in their final semester of the MMA program, Chandra and Hammam saw this as an opportunity to contribute their data analytics expertise to help the public understand the current public health crisis.
Professor Mehmet Gumus, Academic Director of the Master of Management in Analytics program at McGill University, says that Chandra and Hammam’s COVID-19 dashboard is a great initiative.
“Data analytics can have significant impact on flattening the curve by directing the scarce resources in our healthcare system to the right location at the right time,” says Professor Gumus. “This is especially crucial during the time of supply-demand mismatch of resources such as test kits and ICUĚýunits across Montreal and Quebec.
“Given that outbreaks are likely to be transmitted in clusters, using data analytics to contain the community transmission and spread the workload across location and time are essential to avoid overburdening the healthcare system,” adds Professor Gumus.
Follow the data to find the facts
It was important for the two students that the dashboard that provided an ethical snapshot of the developing situation in Canada. Their vision was to create a source of truth to counter the misinformation and panic being disseminated in the wider community.
“We feel that there is a lot of misinformation about COVID-19 on the internet. We need to have systems in place to reduce panic and let people know that the steps the government is taking to fight the virus is seeing improvements,” says Hammam. “The need of the hour is to have a positive mindset.”
In order to tell an accurate story of the situation in Canada, they understood that presenting the bigger picture meant not limiting the dashboard to only tracking cases. The dashboard reports not only on active cases, recoveries, death, and cases per million, but also the number of tests completed.
The media reported on March 23 that Quebec had experienced a significant spike in cases, with the number of confirmed cases in jumping from under 100 to more than 400. “This increase wasn’t reported in the correct manner, with some articles creating a sense of panic in online communities,” says Hammam.
This is where Chandra and Hammam’s decision to report on different metrics proved to be an invaluable addition to the dashboard. While some online publications merely reported on the sudden increase, the COVID-19 dashboard told a different story, with a similar surge in tests completed.
The duo verified their suspicions; the increase was due to new acceptance of tests by hospital centre laboratories that didn’t require confirmation from the Quebec Public Health Laboratory (LSPQ).
Developing the COVID-19 dashboard
Chandra and Hammam have received a lot of positive feedback from academics, colleagues, and peers from across the world. Since launching the website last week, the dashboard continues to receive hundreds of visitors a day.
This growing impact is an impressive feat for a passion project created within 24 hours. The duo developed the dashboard during the initial two-week closure of universities across Quebec.
The two students were familiar with web scraping and Tableau through the MMA course, and knew that defining the solution architecture was a key step in getting this right. Without any financial backing, they were limited to free services—and decided to go with Google Colab, Google Drive, and Tableau Public to bring the project to life.
When they were developing the dashboard, Chandra and Hammam wanted to stay away from any predictive analytics or machine learning models. “We needed to focus on descriptive analytics and avoid creating a complex dashboard that would dilute the main message,” says Chandra. “We needed to place the integrity of the analytics profession above our own interests.”
With several data sources available, the duo opted to use Wikipedia. Wikipedia regularly updates its COVID-19 data from trusted sources like the official provincial websites and major news outlets, so, for Chandra and Hammam, this was an easy choice.
“We can’t achieve 100 per cent automation, because that can only happen when processes are constant or cyclical. In this case, the situation and data source format is changing every day. We fix our codes every three to four days and have set the manual refresh process to less than 10 minutes.”
“We are trying to inform the public about the situation in an objective matter. This will serve as a preliminary source of information,” says Hammam.
Visit the McGill Desautels Master of Management in Analytics website to learn more about the program.
Ěý
Venkatesh Chandra
ĚýRamy Hammam
Ěý
