This project analyzes public sentiment and discussions about the passing of Queen Elizabeth II through a dataset of tweets. By employing unsupervised learning methods like K-Means clustering and Latent Dirichlet Allocation (LDA), sentiment analysis tools (VADER, TextBlob), and supervised learning with logistic regression, we explored global emotional responses and societal reflections on this historic event. Key findings include the identification of prevalent topics, sentiment trends, and the effectiveness of heuristic-based supervised learning in the absence of labeled data. Challenges in dataset quality and cluster coherence were addressed, providing insights into the complexities of analyzing social media data.
Dataset from Kaggle
https://www.kaggle.com/datasets/aneeshtickoo/tweets-after-queen-elizabeth-iis-death