Chatbot and Voicebot Feature for RamiBot Using Deep Learning Sequential Model for Intent Classification

This paper introduces RamiBot, a robot concierge designed specifically to meet the demands of Asia Pacific College and outfitted with both a chatbot and a voicebot module. RamiBot serves as a virtual assistant, providing customers with a wealth of information about academic programs, campus amenities, specific locations, entrance requirements, and important facts about APC. Rami enables speech and text conversations, letting users converse using their preferred method of communication. Artificial intelligence (AI) systems called chatbots are created to simulate human-to-human communication. By providing the relevant information without the need for human participation, they serve to respond to user enquiries and support clients. RamiBot's capabilities are improved with the addition of a voicebot capability as an expansion of chatbot functionality. Users have the option to communicate with RamiBot by speaking, opening up new avenues for ease and accessibility. By building an interaction style that closely resembles natural communication, RamiBot increases user engagement by incorporating speech recognition technologies. This is accomplished by combining machine learning with Natural Language Processing (NLP) strategies, which results in the development of an intent forecasting model. Rami can properly determine the user's message intent thanks to this model and respond to their queries.

I. Introduction

One of the defining features of human progress in the twenty-first century is artificial intelligence (AI), which is also a crucial component of cutting-edge digitization [1]. A number of transformational applications have emerged as a result of the rapid breakthroughs in AI. The development of chatbots, which are intended to mimic human-to-human interactions, is one particularly noteworthy application. Chatbots have been very popular at educational institutions in recent years. They have shown to be useful resources for answering user questions and spreading information without the need for direct human engagement.