GNITS-NaviGlow: College Navigation Chatbot

GNITS-NaviGlow is a cutting-edge chatbot designed to revolutionize campus navigation. This chapter explores the development and implementation of a multilingual chatbot that supports Hindi, Telugu, and English, making it accessible to a diverse user base. The chatbot utilizes advanced speech recognition (ASR) to convert user queries into English for uniform processing, ensuring accuracy and inclusivity. The system employs a logistic regression model, which achieved the highest accuracy compared to other machine learning models, to predict the correct campus location. Once the destination is identified, the chatbot seamlessly integrates with Google Maps to guide users to their desired location. The dataset used for training the model includes 1016 queries categorized into eight distinct campus blocks, such as EEE, CSE, ECE, IT, ADMIN, CANTEEN, HOSTEL, and STATIONERY. The chapter also reviews existing literature on chatbots and their applications in education, highlighting the unique aspects of GNITS-NaviGlow. The results demonstrate the effectiveness of the logistic regression model, with an accuracy of 87.5%, making it the most reliable classifier for this dataset. The chatbot's user-friendly interface allows for both text and speech input, further enhancing its accessibility. This innovative solution not only improves navigation within the campus but also sets a new standard for multilingual chatbot applications in educational settings.