ABSTRACT
Stock market is an important area of research due to its higher earnings. The higher earnings for the stock market also imply higher risks, so a large amount of data generated by the stock market is considered a treasure of knowledge for investors. There are several aspects that affect stock market fluctuations; the most important of them is news data. News data have an influential effect on the investors’ thoughts and beliefs. Using machine learning and textual data processing are considered a significant part of the stock market analysis. Researchers have been concerned with designing a suitable model to predict the future behavior of the stock market to avoid investment risks. A strong relationship has been found between stock news and changes in the stock prices. This study aims to propose a framework for detecting stock market fake news that helps in avoiding higher investment risks and improves stock market prediction accuracy. The study also aims to determine the best combination of machine-learning algorithms that lead to the best performance of the prediction model that is designed according to news sentiment and numeric data analyses. Different experiments have been applied to uncover algorithms that lead to the best performance and raise the prediction accuracy to 92%.
