A new software platform, created by two U of T Engineering alumni, aims to make virtual classrooms more functional by providing real-time feedback and specific insights into how student understanding of mathematics is changing.
Last March, Nived Kollanthara (IndE 1T7 + PEY) was living in New York City, where he volunteered part-time at a shelter, helping kids with their math homework. When the pandemic hit, he realized right away the impact it would have.
“The kids I work with need extra, individual attention to help them succeed in the classroom,” he says. “I was worried about how they would be getting that in a remote environment.”
Kollanthara started talking with teachers he knew and learned that one of the first things that gets lost in virtual learning is real-time feedback on student understanding.
“Assessments and tests can tell you a bit about how a student is doing, but they’re not the whole picture,” he says. “A lot of it comes from seeing how kids are engaging with the content — who’s putting up their hand first, who’s slowing down when certain topics are coming up, things like that.”
Kollanthara started to wonder if it would be possible to build a piece of software that could leverage artificial intelligence and data mining to provide those insights. He contacted his former classmate, Aiden Carnegie (EngSci 1T7 +PEY) to see if he could help.
“During my time at U of T, I worked with two startups to build a product from scratch,” says Carnegie. “This idea caught my interest because of the opportunity to help students learn, and to provide teachers with tools that can help not only during this pandemic, but afterward as well.”
Within a couple of months, the platform, called Kanak, was up and running. The team is currently testing it with a small group of teachers and their students, including some from both Canada and the U.S.
Students log on to Kanak to see a list of assignments provided by their teacher. The multiple choice questions are “gamified” — a correct answer generates fireworks and adds another notch to a student’s “winning streak.”
As students work their way through, Kanak collects information on their responses, such as how long they are taking to answer and what proportion of the questions they get right. Based on this data, Kanak can help the teachers zero on the areas that are causing challenges.
“There are a handful of platforms that provide teachers with real-time feedback,” says Kollanthara. “What differentiates Kanak is the use of deep learning to provide specific insights. For example, being able to determine that a give student takes longer to add fractions because of a lack of understanding of equivalent fractions.”
“I think Kanak is a useful tool because of its ability to personalize the learning experience to meet the individual needs of each student,” says Marissa Sansalone, a teacher with the Toronto District School Board.
Sansalone teaches Grades 1 to 4, with an especially strong focus on math. She heard about Kanak through a friend who shares her interest in STEM (Science, Technology, Engineering, and Mathematics) and the use of technology in the classroom.
“I have been using Kanak outside of the class currently as a practice tool with students,” she says. “I like the insight it provides, such as the warnings that a student is not ready yet to move on to the next concept. In today’s world, tools that enhance virtual learning are essential to student success.”
Over the next few months, Kanak will continue to refine the platform based on the feedback they are getting. By January, they hope to be able to expand their testing cohort to a group of about 20 teachers and their classrooms.
“Our top priority now is getting this into the hands of the teachers and students that need it,” he says. “If they see value in it, the next step will be to identify champions who can help bring this into school boards or other educational environments.”
-This story was originally published on the University of Toronto’s Faculty of Applied Science and Engineering News Site on October 8, 2020 by Tyler Irving