INTRODUCTION 

If Artificial Intelligence (AI) was still a myth for many education establishments, the recent COVID-19 crisis has accelerated its adoption. We know that AI has been introduced early in the educational system since its rise in 1980s. Since then, machine learning (ML) and AI have contributed to altering schooling tools, institutions, and learning styles, and revolutionizing the educational system. According to a research report by Technavio, AI in the US education sector is expected to grow by 47.5% during the period 2017-2021. Its impact has been felt across institutions from lowest to higher education levels.

We invite you to explore further the potential and current application of AI in the education sector:

Automation of administrative tasks

Administrative tasks generally account for significant time spent by teachers and schools, and require tremendous efforts to organize. By automating the process, AI can save time for both teachers and administrators and help them focus more on teaching, tutoring or planning tasks.

Automated grading is a specialized AI-based computer program which can assist with grading activities, assess students’ knowledge, analyze their answers, and provide personalized responses to students. Although machines have previously been able to assess multiple-choice tests, AI takes a step further to grade written answers and regular essays. Teachers can also benefit from AI through its ability to handle routine paperwork, logistics issues, and other personnel issues. By this means, teachers can save energy and spend more quality time  focusing on essential tasks important to students’ development.

Additionally, by using intelligent assistants, administrative tasks such as budgeting, course management, purchasing and procurement activities, expense management, and facilities management are more efficiently managed and more cost-effective. For example, AI is used in the modelling of data to forecast the expenses of the operations departments, thus saving costs in the future and improving the overall responsiveness of the department. AI is also applied in administrative processes to reduce human bias, and improve the fairness and quality of student selection. With given criteria, the AI system can ensure credible and fair admissions and ease the process by providing detailed instructions and prompt assistance to the candidates.

Smart content 

AI is being used to create customized textbooks with digitization of the contents and appropriation of learning interfaces. In other words, smart content is introduced in schools at different levels, from elementary to post-secondary education.

Cram101 uses AI to turn a textbook into a smart study guide complete with chapter summaries, flashcards, and practice tests, thus making them more comprehensible and user-friendly. This includes tools such as navigation. Netex, another useful AI interface, allows lecturers to design digital curriculum with online instructor programs integrated with audio and illustrative video. It also provides a personalized learning cloud platform in which students can customize the learning systems, receive real-time assessment, and stream video conferences, online learning, and other virtual courses.

Personalized & adaptive learning

As each student has distinctive learning styles and preferred interests, personalized learning with the adoption of AI helps tailor educational programs to the specific needs of the learners. AI adapts the training to the learning pace of students and consistently adjusts the tasks to upgrade learning to the appropriate level.

A more promising area of application of AI is adaptive learning.  This method assesses the knowledge level of students and offers advice to teachers to select suitable learning materials or teaching methods to tailor the learning curricula to the individual’s progress. This is also referred to as an intelligent tutoring system (ITS).

Mika software from Carnegie Learning provides a clear example of an ITS. By using principles of cognitive science, its AI-based software provides personalized tutoring and real-time feedback for students who need more assistance, especially freshmen in their first year of college. iTalk2Learn system 16, developed by Carnegie Mellon University, is another model designed to test the effects of adaptive learning on the effectiveness of the learners. With inputs including quantitative knowledge, cognitive needs, emotional status, and feedback from students, the system applies a learner model to tailor the needs of learners and predict learning outcomes. 17ZuoYe (in Chinese means “homework together”), a digital learning platform that has attracted 60 million users as of February 2018, offers online exercises and assessments for Math and English that teachers can assign to students. Parents are also aware of their kid’s homework and progress.

Virtual facilitators 

The ultimate goal in AI adoption is to create virtual human-like characters who can respond and interact in a natural way, using both verbal and non-verbal communication.

With the help of voice assistants such as Amazon Alexa, Google Home, Apple Siri, and Microsoft Cortana, students receive on-time instructions on the education materials regardless of teaching or non-teaching environment, making learning easy and accessible. Jill Watson, a virtual assistant at the Georgia Institute of Technology, has been assisting thousands of students by giving quick and accurate advice. Not popularly known is that Watson is a computer equipped with an IBM AI system.

A chatbot can help students in finding information related to their campus or to their courses, such as the location of the studying hall, emergency contact numbers, etc. The timely and focused answers greatly assist students in their learning. In addition, it saves costs in printing handy handbooks or designing instruction boards. Such initiatives have been applied in many universities, including Arizona State University.

Concerns over AI in education 

However, there are rising concerns over the applications of AI in education, mostly focused on privacy, data protection laws, and human-machine partnerships.

According to the Family Educational Rights and Privacy Act (FERPA), institutions have to (1) get consent to disclose personally identifiable information and (2) give students the ability to access their information and challenge what they think is incorrect. This can be a challenge in other countries where data protection laws are not strong and individual rights to freedom of information can be easily violated.

Experts also show progress in continually updating features of AI for broader applications of adaptive learning to apply practical intelligence to empower critical emotional connections that drive student performance.

CONCLUSIONS 

Although there are rising concerns that AI might replace educators, it needs to be emphasized that AI serves mainly to assist educators in making the learning process more adaptable, individualized, and effective. This can be done through tailored tools, virtual assistants, smart content, and tutoring systems. Though these technologies come with a price, their application definitely yields outcomes in the long-term. However, there is plenty of room to maximize the benefits of AI in education such as integrating social skills, addressing data and privacy issues, global learning, etc. We look forward to more initiatives and applications of AI in education in the near future all around the world.

by Do Ngoc Thao