Exploring AI-Driven Chatbots for Enhanced Learning Experiences in Dental Education with Deep Learning

Abstract

The integration of chatbots has been extensively explored across various fields. However, the clinical domain, particularly in medical education, presents a critical need for im- plementing AI-powered chatbots. To address this, we developed Dentbot, a custom AI-driven chatbot tailored for a predoctoral implant program in dental education. This initiative aims to leverage technology to improve learning outcomes and streamline teaching processes for students. The development of Dentbot involved several key stages, including data preprocessing, model training, and model evaluation, ultimately resulting in the creation of the chatbot. User testing was conducted within a dental institution to gather feedback on Dentbot's usability and effectiveness in the educational context. We evaluated three deep learning models, LSTM and Bi-LSTM, and BERT for chatbot implementation. LSTM and Bi-LSTM both achieved good accuracy rates of 80% and 89%, respectively. Additionally, our study highlights BERT as the most efficient deep learning algorithm for multiclass intent classification, achieving an im- pressive accuracy rate of 94%. These findings demonstrate that Dentbot's development underscores the growing importance of leveraging AI technologies to address specific educational needs within the healthcare sector by streamlining the teaching process and setting a precedent for innovative solutions in medical training and beyond.

Ähnliche Arbeiten