Partner Spotlight: Eyeread
Julia Rivard-Dexter, CEO and founder of Eyeread, Inc., answered some questions for us regarding Eyeread, Dreamscape, and their work with the EFL team.
Please tell us a little bit about Eyeread.
Eyeread (developer of Dreamscape) is a learning platform that delivers research based digital literacy technologies in the form of video games for all learners.
We support educators in the delivery of ELA programming aligned to curriculum with a goal to help players master important comprehension skills. Our programs are unique because of our adaptive learning technology ensuring the content students are presented adapts to the users unique ability. The platform is adopted because of its engagement for kids leading to exceptionally high user adoption.
Dreamscape is our first digital game on the Eyeread platform. It’s unique multiplayer video game design captures kids’ interest leading to increased motivation for reading and learning. Dreamscape offers a student diagnostic for placement into the reading content, reports showing progress over time and current reading skill and assignments that educators and parents can deliver through the games.
The platform is free to use with optional memberships to unlock additional game assets. Anyone interested can set up a free account here: https://www.playdreamscape.com/
As one of our key data collection partners, can you tell us about your involvement with projects across the Partnership Grant?
We’ve been delighted to work with researchers and community partners on the development of questions we want to try and answer as well as projects that help meet the goals of ensuring literacy for learners. Some examples of project include:
Project: Automatic Question Generation
Overview: Using our datasets which include passage and question data aligned to comprehension skills, the team at University of Alberta lead by Alona Fyshe, Canada CIFAR AI Chair and Bilal Ghanam from the Department of Computing Science have built a model to automatically generate questions that will be evaluated by Eyeread educators for accuracy and alignment to the passage.
Goal: The goal of the project is to have a model which is able to reliably generate questions based on text imputed by educators that assist in the development of questions for assignments for learners.
Success: In 2022, we plan to continue the collaboration by adding the outputs of the model to the Eyeread educator workflows to generate feedback from experts that will continue to inform the model to improve reliability.
Project: Presenting Storybooks Canada African Stories through Dreamscape
Overview: Storybooks Canada and Eyeread have partnered to add African Stories to the content library offered to millions of students in the Dreamscape game. To do this, we have generated comprehension questions based on the stories and the stories and questions are reviewed by an external team of DEI reviewers.
Goal: The goal of the project is to have the DEI review competed by the end of February and for the content to be published for use in March, 2022.
Success: Having more inclusive content available to players and educators through the Dreamscape game moving towards more windows and mirrors in the content for learners.
Project: Error Analysis
Overview: We are working with a team led by Jenny Thomson, Reader in Language and Literacy, Director of Research and Innovation in the Health Sciences School at the University of Sheffield, Carrie Demmans Epp, Assistant Professor, Dept. of Computing Science, Director EdTeKLA Research Group, University of Alberta and Carla Hudson Kam, Professor in Linguistics, Director Language and Learning Lab, University of British Columbia in the very early stages of understanding: “Could computers get better at understanding the types of errors children make and respond more adaptively in the delivery of content?”
Goal: We see the potential to identify specific groups of learners, via error patterns, which could lead to more specific in-game supports/scaffolds, or changes to game design.
Success: Our current goal is to scrutinizing log data to understand sequence, pacing, and error patterns of sample data.
How do these projects relate to Eyeread’s goals?
Our goal is the deliver the best education to all learners in a way they love. These project support this mission by:
1) Working on tools that support educators in the development of engaging and student specific assignments,
2) Improving the inclusiveness of our content to contribute to our goal of having curriculum for all learners that is both mirrors and windows to support the development of comprehension,
3) To learn more from our data to continuously improve the learning engine and how we delivers customized learning to each player in the game.