Predictive Analytics and the Future of Learning
Imagine a training and development learning management system (LMS) that is no longer even called an LMS, but has a more familiar, human-sounding name. Let’s call this LMS Lucy, or Luke, depending on the user’s preference.
Imagine that Lucy greets an employee with a sophisticated, personalized dashboard when they login to start their day. This dashboard not only reminds the employee where they left off on current trainings, but makes suggestions about what to study next, or offers up interesting articles in the employee’s domain, or even suggests colleagues or potential mentors that might be able to assist with whatever problem the employee is working on at that time.
Imagine discussion boards or forums that are set up more like a social network, with better UX design that allows easy integration of videos, photos, and articles in an interest-based, informal way.
If all of this sounds familiar, it should. In the consumer market, systems like Amazon’s Alexa, and Apple’s Siri already do very similar tasks. These systems use predictive analytics and machine learning to gather and analyze user data, and then use that data to provide a better experience to the consumer.
Social media networks like Facebook, Pinterest, YouTube, and Instagram allow users to share their interests, skills, and learning journeys in an informal way. And though the method of posting feels informal, the database of collective knowledge on these platforms is vast and searchable, which is perfect for picking up a new skill at the exact moment it’s needed. The sheer scale of content on these platforms is often made more manageable using predictive data analytics, which will present content to users based on past history and by comparing user behavior with other similar users.
In the field of training and development, most researchers agree that predictive analytics and machine learning will revolutionize data collection for instructional designers, educators, and other developers. Not everyone agrees on how that data might be used to benefit the user, which in this case would be an employee within a training and development setting.
Similarly, most LMS applications already incorporate some form of a dashboard, which is available to many different user types, including learners, developers, instructors, and administrators. In their current form, these dashboards aim to provide simple, easy-to-understand visualizations of the data that is being collected for any given course or user. I would venture to state that, at present, these dashboards are significantly more useful to administrators and course creators than they are to learners.
I predict that in the near future, the integration of targeted learning analytics with the ever-increasing sophistication of dashboards, could develop a learning experience in which machine learning not only tracks user data and uses that data to aid learners, it can also foster meta-cognition by giving alarms, alerts, suggestions, or prompts to complete tasks, engage with other learners, or interact with course content. Not only would the LMS provide these assists, it would also give the reasoning behind them.
For instance, rather than just stopping a video mid-stream to redirect a learner to a discussion board, that redirect may also include a brief explanation of why, such as, “Leaners who participate in discussion or reteach what they have learned are X times more likely to retain that information over longer periods of time.”
Or it might offer data-driven insights to users that the users, themselves, may not have even realized, such as, “We notice you perform better on your assessments when you complete them in the morning. Are you sure you don’t want to sleep on this?” In the future everyone could have their own Lucy as a personal learning assistant.