IMMERSIVE INTELLIGENT TUTORING FOR REMEDIAL LEARNING IN VIRTUAL ENVIRONMENT

• Main Author: Rasim
• Format: Dissertations
• Language: Indonesia
• Subjects: Teknik (Rekayasa, enjinering dan kegiatan berkaitan)
• Online Access: https://digilib.itb.ac.id/gdl/view/56228
• Institution: Institut Teknologi Bandung

This dissertation explores the field of learning technology for remedial learning in a virtual learning environment. Specifically, this dissertation researching Intelligent Tutoring (IT) and Immersion that building the Immersive Intelligent Tutoring (IIT) model. This research includes the development of an IIT architecture that is used to build an IIT system and an IIT test to determine user acceptance and its impact in classroom learning IT is a learning system based on Artificial Intelligence (AI) which leads to personal learning. The main problem in IT is knowing the character of learners and what learning strategies are suitable for each learner who has his own interests, boundaries, and capabilities. This study aims to identify learners' character statically and dynamically using Bayesian Knowledge Tracing (BKT). While learning strategies use remedial learning in the form of exercises using faded worked-out examples. In learning, immersion can increase the effectiveness of learning. Immersion can be done by improving the learning process and creating a pleasant learning environment. This research aims to enhance learners' immersion by providing Immediate Explanation Feedback and building a Virtual Learning Environment (VLE) that has rich presentations, user-friendly interaction techniques, adaptive abilities, and provides collaboration between learners. The development of the IIT model uses System Modeling Language (SysML) through its 4 pillars, namely: Requirements, Structure, Behavior, and Parametrics. This model also applies computational theory and learning theory. The computational theory used is: BKT and genetic algorithm. Whereas learning theory uses the remedial method which consists of faded worked-out examples and Immediate Explanation Feedback. The IIT model uses a client / server architecture that involves Learning Management System (LMS), Simulation Linked Object Oriented Dynamic Learning Environment (SLOODLE), and Open Simulator. While testing the model is done through verification of computing, model validation, and educational testing. Computational verification using Root Mean Square Error (RMSE), model validation using feature analysis and educational testing using a combination of Technology Acceptance Model (TAM) and Hedonic- Motivation System Adoption Model (HMSAM) and statistical methods to test the impact of using models in classroom learning. The contribution of this dissertation is the effectiveness of learning in achieving learning goals which include: first is the use of machine learning in the IIT model on the critical components of IIT, namely the student model and the tutoring model. The second contribution is the IIT model consists of 4 basic components of ITS with the addition of the remedial component to the tutoring component, the immersive component to the tutoring and interface component, and the author component to the domain component. The third contribution is IITS for remedial learning VLE which involves learners in the learning process, delivering teaching materials according to student models. The fourth contribution is the user acceptance testing model using a combination of TAM and HMSAM as well as statistical testing to determine the impact of applying the model on learning. The test results show 83% of learners feel happy with learning. While evaluating the impact on learning outcomes through the Mann-Whitney test with Asymp scores. Sig. = 0.490 shows that the use of this model is significantly different from traditional learning. Whereas based on the path analysis test shows that attribute enjoyment and immersion have a positive influence on learning outcomes. However, this research still leaves a gap to be followed up on future research, such as: 1) There are still many attributes that can be used as parameters to determine learner models such as learner behavior in learning, the length of time learners carry out activities in learning activities or the frequency of students visiting the activity site his favorite; 2) The use of other algorithms in machine learning or artificial intelligence that might increase learning technology support in increasing the effectiveness of learning