YouTube Sentiment Analysis on User Opinions Related to Metaverse Using RNN and LSTM

The metaverse concept has gained significant attention in recent years, with its potential to reshape the way we interact, socialize, and engage with digital environments. As discussions surrounding the metaverse continue to grow, it becomes essential to understand public sentiment and opinions about this emerging technology. In this study, the author proposes an innovative approach to gauge public sentiment analysis using a Recurrent Neural Network (RNN) to analyze the responses of certain individual opinions from YouTube commentary on the metaverse to predict and classify them as someone that reacts positively, negatively, or neutrally towards the metaverse and testing it with F1-Score. The result of this study shows that Recurrent Neural Network Long Short-Term Memory Sentiment Analysis on Metaverse-related opinion produced an F1-Score of 66% with a precision of 66% recall at 67% and an F1-Score of 80% with a precision of 80% and recall at 79% for RNN/LSTM Two-Label Classification.