Sentiment Analysis is the computational treatment of opinion, mood, feeling and subjectivity in words and phrases from collected text.
The graphs explore the relationship between sentiment in Tweets, determined by numerical representations, and S&P 500 Index Volatility changes. Sentiment is determined here by a count of the number of positive and negative trading-related words by minute, positive minus negative, and average sentiment is this value divided by the number of relevant Tweets.
Machine learning was used to investigate the relationship further, resulting in the predictions in the Dashboard.
SPX Volatility and Tweet Sentiment by Day
Choose a date:
The chart below is a financial technical indicator that represents historic value changes for each minute of volatility of the S&P 500 Index for a selected day (in black) mapped with predicted values. Both predictive models - Linear Regression and Regression Tree - are based on the Twitter sentiment rate for the corresponding minutes.