Personalized learning has long been a “holy grail” in education. Ideally, we would love to be able to work with each student to achieve a more personalized level of learning that taps into individual interests, skills and desires.
But doing so can take far more time than we have. Our education system was not designed to accommodate this, and moving this mountain seems far too daunting a task.
Or is it?
The growing interest in artificial intelligence may hold the key to a more personalized learning experience.
Artificial intelligence is an important area of opportunity, says Conrad Tucker, director of the Design Analysis Technology Advancement Lab at Penn State.
“Popular digital assistants perform queries, similar to a search engine. What we need is an interactive dialogue system that learns the needs of a student and remembers prior interactions,” Tucker says. “As we move in this direction we will face a number of concerns, such as data privacy.”
Tucker is collaborating with other AI researchers. “The current efforts to advance AI as an academic tool are quite fragmented,” he says. “We need a more concerted effort.”
Stephen diFilipo, interim higher education CIO at Just Right Strategies, says AI can take data mining to a new level. Many schools are building systems that tap into demographics, grades and other data to identify success factors and challenges, and to provide vital student feedback in near-real time, he says.
“AI has the potential to improve how we can analyze both structured and unstructured data built from disparate sources, such as SIS and LMS data, as well as siloed data from different campuses,” says diFilipo. “Algorithms could be built to identify connections, areas of concern, personalized learning resources, useful contacts and more.”
Where are the solutions?
How far have vendors come with providing AI-powered personal learning tools? IBM appears to be leading the way with its education-focused applications of the Watson AI platform.
Watson Classroom includes a mobile app called Element that lets teachers quickly capture student information to generate suggestions on how best to help each student with targeted classroom support. As of April, at least 14 school districts in the U.S. are in various stages of deploying the solution.
Since Watson is a supercomputer, it’s not too surprising that there are few AI-informed software products that support personalized learning using today’s mainstream, X86-architecture-based networking server farms. While many vendors are working on solutions, widely available commercial products are still on the horizon.
Proceed with caution
Artificial intelligence, like many new technologies, is ripe with positive and negative implications. “We must consider things like protecting privacy as we gather and analyze data on a massive scale,” Tucker says.
Much has been written about the inevitable bias written into algorithms when they are developed, and how that might skew the results of tools that use these algorithms in an unforeseen way.
Subhash Kak, a professor in the School of Electrical and Computer Engineering at The Ohio State University, has written about the potential of artificial intelligence to change the face of higher education. But before that happens, there is the ongoing challenge of helping educators leverage these new technologies.
“We’re definitely not using technologies in the most optimal ways,” Kak says. “Schools need to provide more workshops to introduce applications and train faculty in their effective use.”
As we move ahead in this brave new world, academia and industry will work together to identify and protect against these concerns while optimizing the potential of AI to enhance student learning.
Kelly Walsh is CIO of The College of Westchester in New York.