Intelligent Tutoring Systems using Long Short-Term Memory Networks and Bayesian Knowledge Tracing

Educational systems often deliver uniform coursework and exams to all students, irrespective of their prior knowledge, interests, or learning ability. This absence of personalization can lead to reduced engagement levels and diminished learning outcomes. The adoption Intelligent Tutoring System (ITS) is driven by its recognition that each learner is unique, with distinct strengths, weaknesses, and learning styles. Traditional classrooms typically cannot accommodate these variations effectively, leading to a significant achievement gap among students. Moreover, ITS excels at promoting self-directed learning by providing instant feedback, hints, and tailored assessments, which motivates the learner to take initiative in their learning process. This shift towards self-directed learning not only fosters a sense of autonomy and responsibility but also equips learners with valuable skills such as problem-solving, critical thinking, and resourcefulness. This study proposes an ITS which uses Long Short-Term Memory (LSTM) networks and Bayesian Knowledge Tracing (BKT). Through the seamless integration of LSTM across key modules, the Intelligent Tutoring System offers content-level recommendations, dynamic assessments, and real-time support for learner doubts and questions. The study aims to enhance learner engagement, satisfaction, motivation, and knowledge retention by personalizing learning content to individual needs and preferences.