Student Modeling and Analysis in Adaptive Instructional Systems

Abstract

There is a growing interest in developing and implementing adaptive instructional systems to improve, automate, and personalize student education. A necessary part of any such adaptive instructional system is a student model used to predict or analyze learner behavior and inform adaptation. To help inform researchers in this area, this paper presents a state-of-theart review of 11 years of research (2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018)(2019)(2020)(2021) in student modeling, focusing on learner characteristics, learning indicators, and foundational aspects of dissimilar models. We mainly emphasize increased prediction accuracy when using multidimensional learner data to create multimodal models in real-world adaptive instructional systems. In addition, we discuss challenges inherent in real-world multimodal modeling, such as uncontrolled data collection environments leading to noisy data and data sync issues. Finally, we reinforce our findings and conclusions through an industry case study of an adaptive instructional system. In our study, we verify that adding multiple data modalities increases our model prediction accuracy from 53.3% to 69%. At the same time, the challenges encountered with our real-world case study, including uncontrolled data collection environment with inevitably noisy data, calls for synchronization and noise control strategies for data quality and usability.

Hierarchical Deep Reinforcement Learning With Experience Sharing for Metaverse in Education

“…By soliciting feedback, the NPC gains information about the effectiveness of learning materials and about a learner's difficulties. The collected information will be the basis to create what is called a student/player model [6], [7]. By analyzing the data to determine the reason behind previous learning mistakes or success in the past, this type of intelligence helps gain insights into how learners will behave in the future.…”

“…For example, a systematic approach from the basic structure to models creating physical/virtual space to parallel intelligence for decision-makings is required to provide meaningful and enjoyable learning experiences in Metaverse. As reported in two recent surveys [6], [7] any learning environment with adaptation capability must be equipped with both "eyes" to track learners' actions and a "brain" to make decisions on how to dynamically adjust its elements or provide assistance.…”

MetaEdu: A New Framework for Future Education

“…Beyond methods that focus solely on the AI side of ITSs, there has also been extensive developments in data mining [22], big data [23], and multimodal learning analytics [24] for educational approaches, with both areas showing promise for integration with more advanced AI methods. Methods like generative adversarial networks [25], unsupervised learning [26], and clustering [27] can work with student data to spot trends and make predictions that in turn can be used by AI methods to provide appropriate support.…”

“…To tackle this challenge, early research efforts have been devoted to Intelligent Tutoring Systems (ITS), where AI techniques are used to mimic human tutors that know what and how to teach and whom they teach. As stated in a recent survey [3], the long history of productive research of ITS has resulted in successful applications in various domains, such as education [4], military training [5], and healthcare [6], and more work continues to appear. Early ITSs are often described as "homework helpers", where a set of generalized or specific hints is provided upon learners' requests [7].…”

“…With this in mind, ITSs continue to improve their mathematical student models later on through sensor informatics and machine learning. Rather requiring students to ask relevant questions in the first place, ITSs monitor student behaviors in their learning and identify their individual needs for support [8]. Even so, the inability to engage students in ITSs still limits their wide adoption.…”