On Tuesday November 23rd, 2022, a recent graduate of BSc Electrical and Computer Engineering will be sharing a bit more about his novel approach to an Intelligent tutoring system for solution generation with BERT. Stay tuned folks! Let's show our friend some support!
Description: The COVID-19 pandemic has accelerated the need for these institutions to adapt to online and distance learning methods. Intelligent tutoring systems (ITSs) have become common in modern schools and educational institutions. ITSs use artificial intelligence (AI) to simulate human tutoring and provide
immediate feedback, without a human tutor present. Traditionally, ITSs offer adaptive guidance and
instructions to evaluate students by updating the learner’s model. However, there are few studies of ITSs
that use natural language processing and deep learning techniques to analyse the learner’s question and
provide a direct solution. In this talk, I would present how a machine learning technique for natural language processing called BERT (Bidirectional Encoder Representations from Transformers) can be used to build an ITS that can reason the learner’s question and provide an accurate solution. At the end of the talk, I will demonstrate how the various models were trained and we will test the system with different types of questions.