Integrating artificial intelligence into educational technology presents great opportunities for automated educational systems. These systems could relieve teacher resources and support underperforming students. However, creating systems that are adaptive, scalable, and factually correct is resource intensive. Furthermore, there are many technologies that are prevalent, but lack systematic ways to integrate them into existing educational technologies. Building on reinforcement learning and large language models (LLMs), this paper introduces a multi-agent framework for adding both a reinforcement learning-based tutor and an LLM-driven peer to educational systems. The integrated architecture is unified with a central ontology, acting as a symbolic knowledge base and facilitating data transformation. We also detail a novel windowed experience sharing method for improving reinforcement learning training efficiency when dealing with similar environments and low-data situations. We present our architecture and simulated results to verify the reinforcement learning algorithm as an adaptive tutor, as well as the integration of an LLM-driven peer and educational outcomes from this integration.

In the search for more effective education, there is a widespread effort to develop better approaches to personalize student education. Unassisted, educators often do not have time or resources to personally support every student in a given classroom. Motivated by this issue, and by recent advancements in artificial intelligence, this paper presents a general-purpose framework for personalized student support, applicable to any virtual educational system such as a serious game or an intelligent tutoring system. To fit any educational situation, we apply ontologies for their semantic organization, combining them with data collection considerations and multi-agent reinforcement learning. The result is a modular system that can be adapted to any virtual educational software to provide useful personalized assistance to students.

This paper presents a framework for parallel intelligent education that involves physical and virtual learning for a personalized learning experience.We especially focus on Chat Generative Pre-trained Transformer (ChatGPT) owing to its considerable potential to supplement regular class learning.We address the strengths and weaknesses of learning with ChatGPT.Finally, we discuss the challenges and solutions of the proposed parallel intelligent education with ChatGPT.

This work-in-progress paper discusses the development of an educational game to provide integrated geotechnical engineering education modules that connect theoretical concepts, laboratory testing, field investigation, and engineering design. The game, Earth Trek, is developed based on the design of geothermal piles, which are an innovative and sustainable geotechnical engineering approach to combat climate change. Virtual reality is applied to visualize the field environments, laboratory conditions, and design components for structural simulation. The game uses a combination of storytelling and tasks to engage students with geotechnical concepts in an enjoyable way. With the newly developed game, geotechnical engineering instructors can provide students with exposure to laboratory testing and field environments, improving the quality of geotechnical engineering education. The use of multiphysics enriched mixed reality gaming allows for a visual representation of the connections between theoretical concepts, laboratory testing, field investigation, and engineering design. Additionally, this study discusses the challenges that geotechnical students face when dealing with worldwide concerns such as energy demand, environmental protection, infrastructure sustainability, and hazard reduction. Earth Trek allows students to apply geotechnical engineering knowledge to explore the underground space and the associated geothermal energy to tackle the engineering problems using only their smartphones. Through exploring the virtual environment and completing game tasks, students can obtain different testing tools used for geotechnical experiments, including thermal conductivity measurement and direct shear test. They are also trained to conduct parametric study to explore the influence of boundary conditions on thermal transfer efficiency of the geothermal pile. The key contribution of this work is to illustrate an educational paradigm based on mixed reality, moving towards creative engineering education in geotechnical engineering. The newly developed educational game and Earth Trek are expected to enhance geotechnical engineering education and provide students with an interdisciplinary knowledge to tackle worldwide concerns.

This work-in-progress research-to-practice paper provides ongoing results from the development and testing of a personalized learning system integrated into a serious game. Given limited instructor resources, the use of computerized systems to help tutor students offers a way to provide higher quality education and to improve educational efficacy. Personalized learning systems like the one proposed in this paper offer an accessible solution. Furthermore, by combining such a system with a serious game, students are further engaged in interacting with the system. The proposed learning system combines expert-driven structure and lesson planning with computational intelligence methods and gamification to provide students with a fun and educational experience. As the project is ongoing from past years, numerous design iterations have been made on the system based on feedback from students and classroom observations. Using computational intelligence, the system adaptively provides support to students based on data collected from both their in-game actions and by estimating their emotional state from webcam images. For our evaluation, we focus on student data gathered from in-classroom testing in relevant courses, with both educational efficacy results and student observations. To demonstrate the effect of our proposed system, students in an early electrical engineering course were instructed to interact with the system in place of their standard lab assignments. The system would then measure and help them improve their background knowledge before allowing them to complete the lab assignment. As they played through the game, we observed their interactions with the system to gather insights for future developments, which are presented in this work. Additionally, we demonstrate the system's educational efficacy through early pre-post-test results from students who played the game with and without the personalized learning system integration.

This Research to Practice Work-in-Progress paper discusses a next-generation learning system for K-12 students to educate them on scientific concepts surrounding the human body. Specifically, our gamified learning system is designed to make learning more fun, engaging, and effective through game and experiment elements that align with science and math learning standards. It will also increase systematic problem-solving and algorithmic reasoning for K-12 students. Since the human body can be thought of as a combination of interacting systems, the game also introduces students to computational thinking by introducing internal body functions. To achieve these goals, this project has two components. First, an educational virtual reality game will be built according to the natural human body structure. During the game process, students will experience the same as the human body functions, travel along the blood circulation, help with the heartbeat, and participate in oxygen exchange. While students are playing the game, our gamified adaptive learning system tracks and controls the student's learning progress. As the AI component collects student data and uses this information, our system adjusts game content and addresses possible learner issues to refine the learning curriculum for a faster and more effective learning experience. Second, a series of hands-on activities will be conducted based on the functions of the human body (e.g., developing an artificial heart and experiencing how the heart works). Through this project, an attractive and efficient next-generation learning system will be developed and used to expose K-12 students to this learning system. The education of students will be accomplished in different dimensions through games and hands-on practice, respectively. Additionally, compared to traditional learning methods, our learning system not only increases students' interest in learning but also makes it more personalized compared to the conventional learning process. Likewise, we will refer to the results of self-efficacy surveys administered to students and teachers separately to test their perceptions of their abilities and the new system.

With higher education pushing toward larger class sizes, a large portion of current methodology focuses on one-size-fits-all approaches that can effectively educate a large class. However, when these approaches fail, students can be left behind and fail classes due to simple misunderstandings. Inspired by these issues, this paper proposes a modular reinforcement learning system that can be used in intelligent educational systems to inform personalized student support. Based on a similar method detailed in prior work, this paper proposes experience sharing with tutor agents as a computationally light approach to improve reinforcement learning training speed on the task of student support. We also provide preliminary results obtained from student simulations to demonstrate the effectiveness of the proposed method on reinforcement learning agent performance.

Serious games (SGs) are a practice of growing importance due to their high potential as an educational tool for augmented learning. However, little effort has been devoted to address student learning optimization in an SG from a systematic point of view. This article tackles this challenge by developing a learning-embedded attribute Petri net (LAPN) model to represent game flow and student learning decision-makings. The dynamics of learner behaviors in game are then addressed through the incorporation of learning mechanisms (i.e., reinforcement learning (RL) and random forest classification) into the Petri net model for knowledge reasoning and learning. Finally, an algorithm based on LAPN is proposed, aiming to guide learners to achieve a faster and better solution to problem-solving in game. The benefit of the proposed model and algorithm is then demonstrated in the SG Gridlock.

Serious games (SGs) have emerged in recent years as a key method to augment education and training. SGs allow players to experience new concepts while exploring engaging virtual environments. To further extend the educational merit of SGs, adaptive SGs integrate games with adaptive systems to provide a more supportive education. Typically, these adaptive systems either focus on assisting or engaging the player, or on providing a more human-like tutoring experience. For appropriate adaptation, these games make use of player analytics and player modeling to determine what help to provide, predict future player actions, or otherwise model a player’s mental or physical state. To assist researchers in the development of future adaptive SGs, this article reviews and categorizes both player modeling and game adaptation methods from publications over the past ten years. We offer comparisons of various methods to achieve both modeling and adaptation, as well as comprehensive categories for both aspects. Furthermore, we also offer insights for future areas of study in adaptive SGs, as well as possible directions for SG developers.

Metaverse has gained increasing interest in education, with much of literature focusing on its great potential to enhance both individual and social aspects of learning. However, little work has been done to address the systems and technologies behind providing meaningful Metaverse learning. This article proposes a technical framework to address this research gap, where a hierarchical multiagent reinforcement learning approach with experience sharing is developed to augment the intelligence of nonplayer characters in Metaverse learning for personalization. The utility and benefits of the proposed framework and methodologies are demonstrated in Gridlock, a Metaverse learning game, as well as through extensive simulations.

This Research to Practice Work-in-Progress Paper builds on prior developments of a gamified adaptive tutoring system that automates and personalizes a student’s learning process without instructor intervention. To address the continued expansion of general education, as well as the grand challenge of personalized learning, automated learning systems are becoming common within higher education. Our personalized learning system uses an uses a structured, general-purpose game model that enables us to both track and control student progress through the sections of the game. While students play through a system-integrated game, a back-end AI component adaptively chooses both where the student is directed and what help they receive to optimize their learning. The end result is a fully integrated game system that can measure student performance using integrated tests, leveraging that information to adjust game content, address learner misconceptions, and lead to a faster and more effective learning session. As part of continued research, we present results from comparison testing of our educational game system in tandem with relevant course material.With our preliminary results, we focus on demonstrating the system’s ability to provide appropriate content to players based on expert opinion. We show the educational utility of the game system, demonstrating an increase in student performance post-intervention on relevant content tests. We also show results from self-efficacy surveys administered to students to test their opinion of their own abilities. By sharing our testing and verification, we demonstrate the effectiveness of our intelligent educational game system.

Technological advancement has given education a new definition-parallel intelligent education-resulting in fundamentally new ways of teaching and learning. This article exemplifies an important component of parallel intelligent education-artificial education system in a narrative game environment to offer personalized learning. The system collects data on the player's actions while they play, assessing their concept knowledge via k-nearest-neighbor (kNN) classification, and provides tailored feedback to that student as they play the game. Based on an empirical evaluation, the kNN-based game system is shown to accurately provide players with differentiated instructions to guide them through the learning process based on the estimation of their knowledge levels.