Summary of Intent:
In the aftermath of Facebook's transformation into Meta and the advent of the "Metaverse," discussions surrounding "NFTs," "Web 3.0," and "Cryptocurrencies" have seized the spotlight in public discourse. With Meta's positive projections, social media platforms like Twitter have become a canvas for sentiments ranging from enthusiasm to skepticism and neutrality. Our project specifically targets the realm of cryptocurrency, particularly bitcoin, aiming to dissect sentiment nuances and prevalent themes within cryptocurrency-related tweets.
Central to our approach is sentiment analysis, a machine learning method that interprets data based on emotional tone, typically categorized as positive, negative, or neutral. Given the sentiment-rich landscape of platforms like Twitter, our primary goal is to leverage machine learning to cluster and classify tweets based on sentiment and subsequently extract meaningful topics from these sentiment clusters. The project delves into big Twitter data, addressing pivotal questions such as: How can machine learning effectively categorize and cluster tweets based on sentiment? What valuable insights can be derived from sentiment-specific topics? Our work intersects with various research domains to construct a machine learning model for sentiment-cluster-topic extraction from substantial Twitter datasets.
Data Source: https://www.kaggle.com/datasets/226bc93769b5b28697eb8fc4a107040b3145c01640c91c16bde8424170f2820e/metadata