Stocks uses sentiment analysis to parse curated news articles, custom word lists to return a count of positive and negative words and a backend service that uses multivariant logistic regression with these inputs and price data to predict stock movement in the next day, two week and one month time frames.