Forecasting Stock Price Movements Based on Opinion Mining and Sentiment Analysis: An Application of Support Vector Machine and Twitter Data

Abstract

Today, social media networks are fast and dynamic communication intermediaries that are vital business tools, as well. This study aims to examine the views of those who are involved in Facebook stocks to understand the pattern and opinion about the intended future stock price. Yet another goal of this paper is to create a more accurate forecasting pattern compared to the previous ones. Two datasets are used in this paper; the first contains 1.6 million tweets that have already been emotionally tagged, and the second has all the tweets about Facebook stock in eighty days. We conclude that positive news about a company excites people to have definite opinions about it, which results in encouraging them to buy or keep that specific stock. Also, some news can hurt users' views as most of the time, things get more complicated, and uncertainties make it harder to forecast the direction of stock movement. By using text mining and python programming language, we could create a system to be operable in those situations.