AI in Emergency Remote Learning Environments: Intelligent Tutoring Systems Perspective

Introduction

The Education sector has over time struggled against diverse challenges to provide students with individualize and adaptive teaching and learning settings. In this regard, special issues have attracted the attention of the research community for achievement effective educational tools with personalized, adaptive contents, automatic and individualized feedback among other specific characteristics assigned to learning environments. In the middle of the last century, researchers in the field of artificial intelligence (AI) saw an extraordinary opportunity to create intelligent systems to enrich current educational technology with high capacities, in special, those more similar to human tutors. This way, the path for research into adaptation and personalization was being paved to meet student's learning needs.

According to Mavrikis and Holmes (2019), intelligent learning environments (ILE) is a broad category of digital educational interactive applications equipped with features that enable the provision of personalized, adaptive support to students (either by means of task selection or adaption, or dynamic assistance while students are undertaking a task). The same authors refer two target learning systems, those that pursuit step by step the interactions system-student (hence Intelligent Tutoring Systems) when the student is carried out a close-ended learning activity, and those that give intelligent support in this case more oriented to open-ended or exploratory activities.

Regardless of whether it is close or open-ended design approach, the main role of ILE goes hand in hand with the vision of the education as a pilar for a sustainable future (UNESCO, 2014) which aims at designing cost-effective and at the same time quality teaching and learning opportunities for all students. Moreover, different teaching modalities, namely, in-person teaching (i.e., traditional classroom), blended/hybrid teaching (i.e., in-person plus online), purely online education, and distance education, all are more and more part of the daily life of many educational institutions and students’ preferences. Additionally, the massification and heterogeneity of students are also proof of the real changes in our society. It starts to be very common, students of different cultures, ethnicities, social levels, with disabilities, among other situations, be altogether enrolled at all levels of education.

In addition to the mentioned globalization issues, other particular situations can interfere in the education context. A remarkable example is the COVID-19 pandemic. In addition to force a not in-person learning approach it obliged a rapid change of behaviour on the part of all actors in the education sector. Around that time, the Emergency Remote Learning Environment (ERLE) term aroused triggering intensive research and development by the scientific community.

The appearance of ERLE contributed to clarify what online learning actually means along with other terms that orbit around it, namely, asynchronous and synchronous learning, formal and informal learning, and static and adaptive interactions concerning subject contents, learning activities, learning paths, among others (Greenhow, Graham, & Koehler, 2022; Singh & Thurman, 2019). Regardless of the level of education, it appears to be consensual that ERLE should cope with customized, adaptive, and autonomous learning processes. Customization and adaptation to students’ learning needs, preferences and/or interests are transversal to all students regardless the level of education, however the degree of autonomy has a closer connection to the education level. Higher education students are expected to understand and use information critically, to make their own choices and to pursue their own interests. In this sense, the expected autonomy level of university students is higher comparing to pre-university students. However, educational tools can also be designed to stimulate and challenge curiosity, interest, participation and collaboration, creating a more autonomous learning process (Simões & Faustino, 2019).

The use of intelligent educational tools has proved to have an important role in the technological development of adequate solutions. Intelligent Tutoring Systems (ITS) is a relevant example of an educational tool serving as platform to test both learning theories and artificial intelligence techniques (Alkhatlan & Kalita, 2019; Forbus & Feltovich, 2001; Wenger, 1987).