Financial Sentiment Analysis Through Public Disclosures

In this project I am using the public disclosure documents as an indicator to the stock price change of the following day. I investigate the relation between the words that constitute the report and the outcome of the following day. 
I represent the reports as bag-of-words, which enables Naive Bayes method to calculate weights for each words. Those calculated weights are later to be used as an evaluation tool for each individual report's effect on the change of stock price for the following days.