Sentiment Analysis on ChatGPT Reviews

Abstract

Sentiment analysis, a critical aspect of natural language processing, involves the identification and interpretation of emotions, opinions, and attitudes expressed in text. This review explores the innovative approach of employing ChatGPT, a prominent language model developed by OpenAI, for sentiment analysis tasks. This study employs the Bag-of-Word (BoW) technique to analyze sentiment. The research will also focus on creating models through machine learning methods. The dataset is obtained from the App Store, specifically targeting reviews related to ChatGPT. Filtering ensures only English reviews are considered, resulting in a dataset of 22,000 reviews. A preprocessing approach is applied for data cleaning before utilizing BoW for feature extraction. Subsequently, the study involves deploying, training, and assessing classifiers. Evaluation metrics are employed to measure classifier precision. Among the four classifiers tested, XGBOOST (XGB) excels when combined with BoW features, showcasing the highest accuracy. In the context of Bag-of-Words, the XGBOOST classifier achieves an 84% accuracy rate in data classification.

Ähnliche Arbeiten