Evaluation of Support Vector Machine, Naive Bayes, Decision Tree, and Gradient Boosting Algorithms for Sentiment Analysis on ChatGPT Twitter Dataset

Abstract

ChatGPT is a language model employed to produce text and engage in conversation with users. It serves as a tool for generating text and facilitating interactions in a conversational manner. The model was designed to provide relevant and useful responses based on the context of the ongoing conversation. By the increasing popularity of using ChatGPT, it makes it difficult for users to classify responses about the use of ChatGPT. Therefore, sentiment classification of ChatGPT is carried out. The dataset used is sourced from the kaggle website with a total of 20,000 data. The classification methods used in this research include Support Vector Machine (SVM), Naïve Bayes, Decision Tree, and Gradient Boosting. Through the research results, the Support Vector Machine algorithm had the highest accuracy value with 80% compared to other methods, when the data is divided by a ratio of 90:10. This research is expected to help developers and service providers to improve ChatGPT and understand user responses better.