We are excited to bring Transform 2022 back in-person July 19 and virtually July 20 - 28. Join AI and data leaders for insightful talks and exciting networking opportunities. Register today!
Education is one of the domains where artificial intelligence is showing great promise. But while most efforts have focused on augmenting and enhancing the learning experience, challenges remain in the registration and enrollment domains. Students and learners often struggle to find their way to the right courses and classes, whether because of the confusing variety of courses available on online platforms or the complicated processes for enrolling at higher education institutions.
A handful of startups are poised to solve these challenges through AI tools that provide personalized assistance to students during the registration process. So far, their efforts show positive results.
Helping students find the right opportunities
Online learning hub Udacity made its debut in 2012 when Stanford professors Sebastian Thrun and Peter Norvig decided to make their course “Introduction to Artificial Intelligence” available online.
The platform has since grown to offer dozens of courses and nanodegree programs in various computer science fields. The expansion has helped Udacity serve an increasing number of people. More than 160,000 students worldwide enroll for online courses on Udacity every year. However, the expansion of courses has also introduced new challenges.
New and returning students have to navigate the platform’s vast catalog of courses to find the right fit for their skills and experience. Without consultation and assistance, the process can be challenging and result in course abandonment and lower enrollment rates. Meanwhile, providing personalized counseling to thousands of users requires human resources that would go beyond the company’s capacity and budget.
To meet this challenge, Udacity recently teamed up with Passage AI, a startup that specializes in conversational interfaces. The result of the partnership was an AI-powered chatbot that helps students visiting Udacity’s website to find appropriate courses and enroll with ease.
“Our students are looking for guidance and quick responses to questions while browsing our course list,” says Sam Mazaheri, director of digital marketing at Udacity. “It is important for them to find classes that fit their needs and experience level. A chatbot can provide immediate responses to common questions, giving prospective students confidence that they’re enrolling in the right program at Udacity.”
Passage AI’s chatbot uses natural language processing (NLP), a branch of artificial intelligence that enables software to analyze and respond to the meaning behind human-generated text. NLP helps chatbots move away from rigid command forms and enable their users to interact with them in a comfortable, conversational fashion. This provides students with an experience similar to what they would get from talking to a professor or counselor instead of manually finding their way around the courses.
The chatbot was initially tested with 5 percent of Udacity’s visitors and eventually rolled out to the entire audience. The results, according to Passage AI CEO Ravi Raj, was a 40 percent increase in click-through rates, indicating a decrease in confusion and more students finding the course they’re looking for.
“We see AI-powered chatbots guiding students to find the right course or nanodegree on a site like Udacity or answering frequently asked questions like how much a course costs or how long it would take to complete it,” Raj says. He foresees chatbots eventually helping with career planning and counseling by having an ongoing dialog with students and recommending the right courses of action to accomplish their goals.
Providing individual assistance
The process of applying to college involves several challenging tasks, such as completing financial aid applications, submitting a final high school transcript, obtaining immunizations, accepting student loans, and paying tuition.
The process is especially difficult for students who come from low-income backgrounds or who don’t have a family member who can guide them through the process. Not providing support to students results in many of them failing to matriculate. Meanwhile, every student is faced with their own unique set of challenges, which means generic outreach tools may prove inefficient and cause further confusion.
An alternative would be to provide individual counselor outreach, but it would not scale as the number of students increases. An interesting case study in this regard is Georgia State University, which partnered with AI-powered messaging platform AdmitHub to provide assistance to students in their transition to higher education.
GSU used AdmitHub’s conversational AI system Pounce to automatically reach out to would-be college freshmen based on their individual pain points. Pounce analyzes the university’s data on each student’s progress in accomplishing pre-matriculation tasks to understand where they might be struggling and where they need help. Using this information, it generates personalized text messages to only reach out to students who are in need of assistance instead of sending out generic messages to everyone.
Pounce also uses NLP/NLG technology to provide smooth conversational assistance on general inquiries around the clock, providing assistance at scale. As with all technology that involves deep learning algorithms, Pounce grows smarter as it interacts more and more with students.
What makes Pounce especially interesting is its proactive working model. Instead of requiring students to reach out for guidance, Pounce makes the outreach based on the data it has.
An experimental study found that students who received Pounce outreach completed their required pre-matriculation tasks and enrolled on time at significantly higher rates than those who received the standard assistance. As a result, GSU observed a 21 percent decline in summer melt, the phenomenon of prospective college students’ motivation to attend higher education “melting” away during the summer between the end of high school and the beginning of college.
AI has already proven that it can be a handy tool to provide quality, personalized education. Now, with the help of these and other similar projects, it will assume an active role in putting more students in classes.
Ben Dickson is a software engineer and the founder of TechTalks, a blog that explores the ways technology is solving and creating problems.
VentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Learn more about membership.