¾ÅÐãÖ±²¥

Learning a new language or mastering technical terminology can be daunting. But imagine having a personalized AI tutor that understands how you learn, has comprehensive knowledge of common mistakes and errors learners like you make — and can use that knowledge to help make it easier for you to learn. 

Those are the types of tools and technologies that , an associate professor in the Department of Computing Science, is developing in the research group she leads. 

As she explains, well-developed tools and technologies powered by artificial intelligence have the potential to make a big difference in education by improving student learning and reducing teacher workloads. However, there’s one problem — many of these educational technologies “fail to account for theories of how people learn” or “are divorced from the realities of the educational contexts in which the technologies are meant to be used.” 

Demmans Epp was recently recognized with the 2024 Killam Accelerator Research Award for her research, which bridges that gap between technology and education by looking at how people are really learning and creating tools and solutions to support them so they can learn more effectively. 

The lab’s project is the perfect example. When learners begin studying a new language, they bring context from the languages they already know. The way they structure sentences in their native tongue, for example, might “transfer” into their writing in English. The project offers an effective solution to this issue — by using machine learning and language models, it identifies writing errors and provides personalized feedback to learners through a Google Docs plug-in.

Since language learning is about sound as much as syntax, they also created , an arcade-inspired game that teaches learners Plains Cree. As the title implies, SoundHunters focuses on the sounds of the language, not just how it’s written. A version teaching , a traditional Métis language, has also been tested.

“What AI is really good at is getting a sense of what a person knows and doesn’t know, and then changing what they’re being given from a learning content or activity perspective so they learn more,” says Demmans Epp.

She often shares her expertise with international learners trying to navigate the world. Recently, she was in Japan doing field research, working with students and academic advisers to co-design tools that help with self-directed language learning. 

“It’s about training them to track their own learning, decide what they need to do, create a strategy and plan, and then work towards achieving the goals they’ve set.” 

While language learning is a big focus of her lab, Demmans Epp also applies her interdisciplinary lens to learning things that are so complex they almost seem like a new language. The Tier 3 CS Vocabulary Project, for example, uses natural language processing to create a list of advanced computer science vocabulary that gives students or instructors the field-specific knowledge they need. 

In addition to streamlining the way learners acquire knowledge, whether a new language or a new set of field-specific terms, Demmans Epp’s lab also seeks to better understand how people learn. 

“We’re using different sensors and other ways of figuring out how much effort people are putting in, the different types of effort, and how that affects their learning.”

One project, for example, uses eye-gaze analytics to understand how learners focus their attention and seek information. Another draws on classification models to determine how high school students look for information during online reading tasks. 

Along with pursuing topics that intrigue her as a researcher, she often goes straight to the source — learners themselves — for inspiration on what future questions her lab should explore. 

“We’ll often elicit experiential information from learners and then target what we’re working on based on that, or we’ll have learners approach us and propose things,” she explains. “We’ll actually involve them in the design and selection of the problem.”

As for what she’s proudest of in terms of her research, Demmans Epp immediately thinks of the people, not the projects, in her lab. 

“I have a really great group of students who are really supportive of each other,” she says. “Sometimes they’ll come back and help each other, even if they’ve graduated.”

That focus means the EdTeKLA lab is also committed to nurturing the next generation of learners, taking on summer students as part of the Faculty of Science‘s WISEST Summer Research Program and, in the past, the computing science department’s High School Internship Program. These programs give graduate students in her lab a chance to develop their mentorship and leadership skills, Demmans Epp notes, in addition to offering young scientists a glimpse into a subject that interests them.