Methods and systems for predicting market behavior based on news and sentiment analysis

The present invention provides a method, system and software that provide a predictive model responsive to the correlation of news articles to stock price movement. The invention analyzes the derivative or ratio of events to drive predictions in a responsive manner. The invention considers derivatives or ratios of news meta-data within a small window in the past relative to a larger window of news items in the past. The invention may use a sentiment engine and apply the derivative of sentiment to predict volatility and/or trend direction of price of a security. The invention may look to the content, context, and derivative of sentiment to weigh news stories according to a predetermined taxonomy factoring in recency, criticality, repeatedness, trustworthiness, etc. to predict stock price behavior. Also, the invention may be used to forecast events given stock price movement and news to predict an impending story or release of significance.