Comparative Analysis of Google Dialogflow and Rule- Based NLTK Chatbots for Application FAQ

Abstract

This study presents a comparative analysis of two chatbot frameworks, Google Dialogflow and rule-based NLTK (Natural Language Toolkit), for the development of chatbots to handle frequently asked questions (FAQ) in applications. The study focuses on Blender, a popular 3D modeling software, as a case study. Ten testing questions were used to evaluate the chatbots' accuracy, precision, recall, and F1-score. The results showed that Dialogflow achieved an accuracy of 80%, precision of 80%, recall of 100%, and an F1-score of 88.9%. In contrast, the rule- based NLTK chatbot achieved an accuracy of 60%, precision of 66.7%, recall of 80%, and an F1-score of 72.8%. The study concluded that Dialogflow is a more effective and reliable chatbot for handling Blender FAQs due to its ability to retrieve relevant information from a large knowledge base and its use of machine learning algorithms to improve its performance over time. However, the rule-based NLTK chatbot may still be useful in certain situations where a more simple and customizable chatbot is required.