Smart Learning Using Autonomous Chatbot Based on NLP Techniques

Abstract

Artificial Intelligence (AI) technologies have shown exponential growth in different areas. Many companies and organizations adopt AI in their workflow to perform tasks efficacy and efficiency. Schools and universities were not an exception. They integrate several AI solutions to help students in their learning processes. AI chatbots are among the AI technologies widely used in the education domain. They understand human language and interact correctly with it through two main components: (1) Natural Language Processing techniques to understand users’ requests and generate the appropriate answers. (2) Knowledge Base (KB) to store and centralize all chatbot knowledge. Chatbot’s KB is the brain behind the chatbot. It is the central component behind any AI Chatbot. However, several researchers still construct the chatbot’s KB manually which consumes time, energy, and cost. In our paper, we propose an autonomous solution to collect a chatbot KB, preprocess it, and store it through the Python programming language. Specifically, to demonstrate the feasibility of our solution, we focus on creating a solid KB for an educational chatbot that answers students’ questions to offload teachers and automate repetitive tasks. In our paper we make four main contributions: (1) Collect the chatbot’s KB through our autonomous solution. (2) Preprocess the chatbot’s KB and makes it clean to use. (3) Understand students’ requests by applying four main techniques: Latent Dirichlet Allocation (LDA), Term Frequency-Inverse Document Frequency (TF-IDF), BERTOPIC and KEYBERT. (4) Develop an AI chatbot to explore the KB and satisfy students’ needs.