Developing an Adaptive Educational Chatbot for Personalized SQL Tutoring

Today chatbots are the most popular assistant tool we come across while surfing any website, browser, application. Chatbots are multi-functional interactive computer programs that can help a user to do a task or chat by using the natural language dialogue. The use of these bots is made for many purposes by the users-including the education. Therefore, this study purposed an educational chatbot for the purpose of tutoring students in the natural language by acting as a human-like teaching assistant. This bot may provide great help to a student in carrying out their exercise, solving problems and modifying the learning pace to meet the requirements of every student separately. Structured Query Language (SQL) was given to students as per their capabilities and customized solutions to their problems. For the purpose of this study-the Rasa Open-Source Framework was utilized to test the framework. Two chatbots were tested-one with the trained model with RASA suggested pipeline and the other trained in the Spacy pre-trained model. The study reflected that these models produced results with similar accuracy, response-time, F -1 score, and precision. However, the chatbots of SQL which are trained with the RASA recommended pipeline resulted in performing better with a 0.01171 p-value in the intent classification.