Maryland university launches AI-powered education tech hub with cross-country partners

The University of Maryland has launched the Center for Educational Data Science and Innovation (EDSI), taking a research-first approach to developing AI-powered educational technology. Unlike many edtech startups rushing to market with AI tools, EDSI prioritizes building ethical infrastructure and conducting rigorous research before deploying solutions. This deliberate strategy represents a significant counterbalance in a field increasingly dominated by rapid AI tool development, emphasizing evidence-based innovation over speed to market.

The big picture: EDSI aims to address persistent educational challenges through AI while prioritizing responsible development and research validation before releasing new tools.

• The center focuses on building infrastructure that enables ethical AI innovation rather than rushing to produce more AI applications without proper research foundations.
• Director Jing Liu emphasizes that this research-first approach fundamentally differentiates EDSI from other AI education centers and institutions.

Key details: The center is continuing development of M-Powering Teachers, an app created in 2023 that provides math educators with non-evaluative feedback on classroom recordings.

• The tool uses natural language processing to analyze transcripts of classroom sessions, assessing elements like student engagement, question strategies, and mathematical discourse.
• M-Powering Teachers provides specific prompts for teacher reflection, helping educators improve without the time constraints of traditional coaching models.

Research foundation: Studies published by the EDSI team demonstrate the effectiveness of automated feedback in improving instructional quality and student outcomes.

• Research shows particular benefits for online learning environments where traditional observation may be challenging.
• Recent papers by Liu and colleagues reveal AI-generated feedback can support teachers in making data-informed improvements to their practice.

Behind the numbers: EDSI is creating a comprehensive database of classroom materials to train more sophisticated educational AI tools with holistic classroom understanding.

• The database includes classroom transcripts, audio, video, student demographic data, achievement metrics, and human annotations on teaching practices.
• This robust dataset aims to improve tools like M-Powering Teachers by providing a more complete view of classroom dynamics.

What they’re saying: “It was important to establish a center that would both support AI use in improving teaching and learning and prepare students for an AI-infused future,” Liu said in a public statement.

• Liu emphasized the center’s focus on addressing critical educational challenges like “declining literacy rates, chronic absenteeism and diverse learning styles” through new technological possibilities.
• “All the stakeholders need to be at the table to do this work really well,” Liu added, highlighting the importance of collaboration.

Why this matters: The approach creates a counterbalance to the often rushed development cycle of educational technology, potentially leading to more effective and equitable AI tools.

• By prioritizing research and ethical considerations before deployment, EDSI’s model could reduce potential harms from untested AI systems in sensitive educational environments.

Collaborative approach: EDSI has established partnerships with diverse school districts and organizations to ensure research and innovation includes varied perspectives.

• Partner districts include Guilford County Schools in North Carolina and San Francisco Unified School District in California.
• The center also collaborates with the Research Partnership for Professional Learning and serves as a hub for cross-disciplinary work within the University of Maryland.