Cryptocurrency Sentiment Analysis using Bidirectional Transformation
Himanshu Dwivedi · 2023
This paper predicts sentiments of crypto currency news articles using BERT (Bidirectional Encoder Representation) model, as there is a lack of research in crypto currency price prediction using natural language processing. The text data obtained is unlabeled and it is labelled using a parsimonious rule-based model and then BERT is used to dassify news sentiment as “Positive”, “Negative” or “Neutral” which may be helpful in reading cryptocurrency market movement.