The Prediction of Intraday Stock Market Movements in Developed & Emerging Markets using Sentiment and Emotions from Twitter

Paper investigates the predictability of stock market movements using text data related to stock markets extracted from the social media platform Twitter. We use high-frequency intraday data rather than daily data and analyse and compare results for both emerging and developed markets. To this end, the study uses three different Machine Learning Classification Algorithms: the Naïve Bayes, K-Nearest Neighbours and the Support Vector Machine algorithms. Several model metrics such as Precision, Recall, Specificity and the F1-Score are also used. Lastly, we use K-Fold Cross-Validation to validate our machine learning models’ results and applicability to unseen data. The predictability of the market movements is estimated first by using only sentiment and then using a combination of sentiment and emotions. Our results indicate that investor sentiment and emotions derived from stock-market related tweets are significant predictors of stock market movements. This model does not only give good results in developed markets but also emerging markets.