Education Today Education Today Magazine
Artificial intelligence and machine learning can identify students who are struggling, enable early interventions, and create tailor-made learning pathways. By Tony Lteif, EMEA Head of Education at AWS
AI and ML are transforming the field of education with powerful insights about the way students learn best. They enable us to monitor an individual’s learning progress, develop intelligent tutoring systems (ITS) that can answer questions and give feedback on assignments when teachers are not available, and gather data through continuous assessment to gain meaningful insights about learners. From AI-powered assessment tools to learning management systems, which allow educators to create personalized learning experiences, AI and ML offer educational institutions powerful tools to ensure the success of every student.
Most importantly, advances in AI and ML have powered new tools that can identify students that are at risk of failing a course, dropping out, or who are exhibiting behavioral cues that indicate the need for extra support. Educational institutions traditionally used prior performance warning systems, which flag existing issues with academic performance, attendance, and behavior to identify the students that are most at risk. Major limitations of traditional warning systems are that they present an incomplete image of student knowledge improvement areas, offer a static educational approach that does not adapt to varying learning styles, and most importantly, that interventions often occur too late to make a meaningful impact on educational outcomes. AI and ML enable schools to quickly and accurately process large volumes of data that take into account prior academic performance and much more: the U.S. Department of Education found that 16-44% of schools surveyed also consider complex out-of-school factors like campus infrastructure, gender, family structure, and socioeconomic background. Highly-customizable warning systems using these technologies offer more than just a risk rating, and can be developed to create predictive timelines: knowing the projected number of days before a student may drop out of school or fail a class allows educators to provide critical interventions earlier.
AI-powered diagnostic testing reveals knowledge gaps
Many edtechs are pioneering AI and ML technologies to offer earlier interventions to students in need. Alef Education’s intelligent learning platform uses adaptive diagnostic testing and an exam prediction model to alert educators about students who are struggling. The diagnostic test questions begin at a randomized difficulty level and then adapt to students’ responses in real time to evaluate their knowledge across multiple expertise areas, setting aside assumptions about a student’s skill, grade, and language levels. After moving the platform to the cloud, Alef Education uses services like Amazon SageMaker, which creates machine learning models to better understand each student’s learning behavior, and Amazon RedShift, which allows Alef to process over 100 million unique data points that reveal nuances about each student’s abilities and generate intelligent recommendations for teachers to support students effectively. More insights can be gleaned from student feedback about test questions: Alef Education designed a user feedback mechanism where ratings about question difficulty can reveal patterns that indicate underlying issues with reading or problem-solving. The AI-powered prediction model tracks data including user activity, completed courses, and feedback to identify students at risk of failing and share those warnings with teachers so that they can intervene early and help students get back on track.
Engage struggling students with tailored learning paths
AI and ML enable students to benefit from a high degree of personalization, connecting users to the right materials at the right time in their educational journey. Madrid-based digital education organization Odilo utilizes AI to provide tailor-made learning experiences for users from schools, universities, governments, and enterprises. Its vast e-learning content catalog connects users all over the world to more than four million educational resources including books, classes, and audio/video content in 43 languages. The company’s Unlimited Learning Ecosystems, intelligent content platforms, analyze data about each individual to provide recommendations and learning plans that are constantly adapting in real time. “We are revolutionizing the way we learn by heavily personalizing the experience for each person. From intelligent profiling to customized learning pathways, it empowers users with AI-driven content selection, tailoring formats, themes, levels, and language,” explains Fátima Bigeriego, the company’s strategic partnership director. “Odilo has moved from traditional standardized learning to focus on the individuality and particular way of learning of each user.” Analyzing data to create a seamless experience for over 170 million users is no small feat, and Odilo utilizes over 40 cloud services including AWS Personalize, an ML service that analyzes user interactions like content views and purchase history to create a highly personalized learning experience that adapts in real time to the needs of each group and individual. The Community of Madrid implemented Odilo’s learning ecosystem to boost youth literacy and learning skills, and within three years, users accessed 57,300 different educational materials over 2.5 million total learning hours. Odilo created over 4,500 individualized reading plans that were tailored to each child’s interests and knowledge level for a more engaging experience. “We envision education in a way that goes beyond imposition and regulations, where users are empowered to know, desire, and choose to learn mainly because the entire ecosystem adapts to them,” Bigeriego says. “In this journey, the cloud is indispensable, as the massive datasets required to make this a reality need storage and other services,” including Amazon Sagemaker, which allows Odilo to classify its massive content catalog with greater precision and make accurate recommendations. “In this regard, AWS has been our steadfast partner from the beginning, supporting us day by day in reinforcing our goal.”
Technology that supports students’ mental wellbeing
Some edutechs are using AI and ML to empower students to take better care of themselves during their educational journey. UniWellBeing created a digital platform that students can use independently with interactive tools that reduce stress, enhance focus, and improve mood – right from their smartphone. Studies have shown that university students experience significantly more psychological distress than the average person: 42% of university students in the UK reported emotional or mental health issues that require professional assistance, and university students are 3 to 4 times more likely to experience a high level of psychological stress than non-students of the same age group. However, they can be hesitant to use university-led mental and emotional health services when they are struggling, with two-thirds of students seeking no assistance or treatment. UniWellBeing’s wellness programs are designed with the student journey in mind: the platform uses data to identify difficult transitional periods for students and deliver the content and tools they need at the time that they need it, from calming breathing exercises to anxiety-busting interactive content that builds better habits. Students can discreetly report on their daily moods, productivity, and struggles with procrastination and anxiety. The platform’s powerful analytics tools ensure that students immediately receive the content and support they need, and give university administrators valuable insights about wellbeing behaviors and outcomes for their student body. Mental and emotional wellbeing is an incredibly sensitive subject, so gaining the trust of universities and students was a top priority. The platform was built using proven methods in behavioral science and mental health outreach, and to protect sensitive student health data, UniWellBeing uses AWS cloud solutions to stay compliant with privacy regulations and ensure that data stays completely private and secure through Amazon Virtual Private Cloud, which hosts AWS resources in an isolated virtual network that offers security features like advanced network access control, without the need for VPNs or physical datacenters. In order for UniWellBeing to support students around the world in Australia, the UK, and the US, they needed to comply with local data privacy laws and obtain security certifications such as ISO 27001, which ensures the secure handling of highly personal student data. “We migrated to AWS because it was the simplest and quickest route to achieve ISO 27001, whilst also supporting our customers globally and removing barriers in order to enter new territories,” UniWellBeing CEO Hugh Griffiths explains.
Unlocking new ways to reach students in need
With continuously-growing capabilities to process and interpret data, educational solutions using AI and ML offer deeper insights about learners, pinpointing what practices or processes aren’t working and which are guiding students towards success. AI and ML offer tremendous advantages in the field of education, and these technologies are transforming the way that institutions teach and students learn: imagine a class where lectures are transcribed and translated in real time so that students receive complete notes from the lesson, can review automatically-generated summaries of important concepts, and instantly create quizzes based on the course materials to test their knowledge. AI tools like automated test assessment reduce tedious administrative tasks for educators and free up valuable time that can be spent with the students who need it most. AI and ML can enable a tailored content experience, identify knowledge gaps, and deliver critical resources to students when needed. But it can also do so much more: Intelligent tutoring offers round-the-clock support for learning challenging subjects. AI-powered translation, transliteration, and rich image description tools make learning more accessible to all. Other tools offer content and student knowledge level summarization, illustration generation, foreign language assessment, “talking textbooks,” and content adaptation according to grade level. These technologies offer a lifeline to students in need, to ensure that no one is left behind.
