An Approach to Detect Sentiment on Public Opinions towards ChatGPT

The usage of Chatbots developed swiftly in numerous fields in current years, together with Advertising, Assisting Systems, Training, Health Care, Cultural Heritage, and Amusement. ChatGPT is a natural language processing tool pushed with the aid of AI generation that permits you to have human-like conversations and much more with the chatbot. The feelings and mind of those who first experienced ChatGPT are valuable feedback to evaluate the achievements and positive and negative factors of this era. In this study, sentiment analysis is used to apprehend people’s enthusiasm about ChatGPT. We have used a collection of English opinions towards ChatGPT. Opinions are collected from different social platforms such as Twitter, YouTube, Facebook, etc. After using ML and DL algorithms such as SVM, DT, LR, RF, NB, GB, ANN, LSTM, and Bi-LSTM, we presented comparisons between different types of ML and DL models for sentiment analysis. We have adopted the feature extraction technique and the SMOTE method for data management and tested the performance of our proposed methods on the dataset. Our experiments indicate that SVM and LSTM show superior results to other models in different categories. And we can understand people’s sentiments about ChatGPT. The findings suggest that a large portion of the initial users of ChatGPT found the experience to achieve success and become satisfied with ChatGPT. However, negative feelings consisting of worry and challenge had been additionally located in some user.