Topic modeling and sentiment/emotion analysis of Elon Musk's all time Tweets with an academic paper. LDA, Top2Vec, BERT.
Data collection:
Tweets (including replies) were scraped using the library: SNScrape
Topic modeling:
Gensim (for topic modeling using LDA/ CoherenceModel for calculating model coherence), pyLDAvis (for visualizing topics), Top2Vec, Spacy, NLTK, Seaborn.
For sentiment and emotion analysis of all Tweets:
A RoBERTa model (by Hugging Face for sentiment/emotion analysis: pysentimiento, which is trained on a huge corpora of millions Tweets, was perfect for this task.