Editor’s note: This article was originally published in edtechdigest on January 18, 2017.
Even a few years ago, most of us would have never imagined a teacher speaking into a smartphone to ask a nuanced question about a student’s recent performance and getting a clear, sophisticated answer in response. Now, given our familiarity and comfort with voice-interactive services—Amazon has over 5.1 million Alexa units in use, more than half of iPhone users engage with Siri on a regular basis and Google Home sold out of a number of stores this holiday season—such a future seems well within reach.
For all the hype in recent years about learning analytics revolutionizing teaching and student success, we haven’t made nearly enough progress.
Artificial intelligence (AI) applications like these come in various forms, such as machine learning, natural language processing (NLP) and deep learning. Their impact is already being felt in many industries and in our daily lives. Our computers are increasingly capable of making sense of vast and disparate data, deriving insight and helping us solve problems quicker by engaging and learning rather than merely responding to a programmer’s commands. A recent McKinsey analysis provides compelling evidence that advancements in AI drive personalized service delivery, and improve optimization of resources in a variety of sectors, including marketing, healthcare, manufacturing, and telecommunications.
AI for Education
In education, we are just beginning to leverage these tools. As educators, we face an ongoing challenge and responsibility to improve the quality and impact of how and what we teach our students. Digital technologies have helped expand access to content and affordable prices, and have spurred experimentation with new teaching and learning models. Advances in AI are poised to accelerate this progress, presenting opportunities for a more personalized approach to learning that meets the needs and goals of every student. It also allows for the scaling of effective coaching practices, and resources that provide the support and guidance students need to persist and thrive.
The implications are exciting. We have an abundance of data from student information and learning management systems, as well as plenty of other tools. We do not yet make sufficient use of those data resources. We also are not reliably capturing meaningful data that can help us understand, not just when students are struggling, but preemptively target strategies that support each individual’s learning needs. Written and verbal student comments in discussion groups and to coaches and advisors are difficult to analyze at scale, yet they offer some of the most important insight into the risk factors that derail student progress.
A number of recent studies have taken up this challenge, analyzing sentiment data from students in MOOCs. Applying natural language processing technologies to parse, messy and large amounts of raw data from student comments, discussion groups and quizzes, the researchers were able to better predict which students were at-risk of not completing the course than using more traditional demographic and login data. These results show great promise. Faster and more accurate analysis means instructors and advisors can get important insights to help students, and tailor interventions and support to meet individual needs in real-time.
Let Teachers Teach
This last point is crucial. The promise of AI in education is not that it will replace teachers and advisors. Rather, it will empower them. For all the hype in recent years about learning analytics revolutionizing teaching and student success, we haven’t made nearly enough progress. The data too often remains inaccessible, underwhelming or it simply arrives too late to have an impact. However, timely accessible insight from data can help educators and students solve their challenges and make better decisions. Using data more effectively, and putting it in the right hands at the right time, makes genuine student-centered learning feasible.
To prepare for these possibilities, teaching and learning platforms must do a more effective job collecting the full range of learning data. That is our mission at my company, and our learning relationship management platform is designed to empower educators by capturing that data and making it more accessible and usable.
We see AI as a key part of that goal. From natural language processing of a broader range of learning data, to more robust and accurate predictive capabilities, to deep learning that enables coaches and instructors to query student learning data in a more natural and intuitive way, in 2017, we are focused on exploring and applying these emerging technologies, validating their early promise and tackling the privacy concerns they may create.
Data has long held great promise for improving student outcomes. By embracing these advances in AI to better serve educators as they support individual learners, we can fulfill that promise.