Sentiment Analysis Based Fuzzy Decision Platform for the Saudi Stock Market

Abstract

Investors in the Saudi stock market are used to seeking advice from online resources such as Twitter and discussion forums. In this paper, we introduced some components to enhance decision making using sentiment analysis and simple fuzzy decision. We built two corpora and one lexicon manually, and we did analysis on them using both corpus-based approach, and semantical and lexical based approach. The best model was selected to be the base of a fuzzy decision mechanism provided for investors. We mentioned several performance metrics, but the main metric to count on was recall. The best model was the rule-based approach with minimum and maximum recall of 69% and 96% respectively as we go through different data sets, types, and sizes.