Predicting companies stock price direction by using sentiment analysis of news articles
Predicting companies stock price direction by using sentiment analysis of news articles

Summary/Abstract: The purpose of this paper is to predict the stock market direction by analyzing the newest financial news downloaded from reliable sources, targeting the big companies. The focus will be on lexicon-based and machine learning models, in order to calculate the polarity and subjectivity of news and explore the relationship between the specific features of companies and stock prices movement. Some of the trained models work with aggregated news for all companies. Furthermore, for more detailed overview there are models trained for specific companies. The main idea is to observe the impact of different types of news on stock prices and resolve the connection between the calculated sentiment and the rise and fall of stock prices. We created a platform for following, downloading, processing and storing financial news articles in real time with possibility to analyze historical news. The processed data is visualized with graphs. This platform provides real-time news articles and stock prices for a chosen company and historical data, all of which help in gaining insight into the aftermath of publishing certain article on company’s stock price.

• Details
• Contents