Analisis Sentimen Opini Publik Terhadap Chatgpt di Twitter Menggunakan Metode Naive Bayes

Abstract

The presence of ChatGPT, which is based on Artificial Intelligence (AI), has attracted public attention and become a polemic. This application is very helpful for users in finding information, translating languages, and generating text similar to human speech. OpenAI's decision to launch the ChatGPT Plus service is related to its monetization strategy and maintaining user interest. In this regard, a study was conducted using the Naïve Bayes method to classify the sentiment of the Twitter community towards ChatGPT in the period 30/11/2022 to 31/12/2022. This study took 5001 tweet data from the available dataset, and used TF-IDF feature extraction and BOW feature extraction, with data division into 70% training data and 30% test data. This research gets the highest accuracy value of 59% using BOW feature extraction.