Sentiment analysis as a factor included in the forecasts of price changes in the Stock Exchange

The aim of the study was to show that sentiment analysis can improve the quality of stock price prediction on the stock exchange. In the study conducted the author proposes an experiment procedure, which, with the use of sentiment analysis, allows to recognize that the introduction of the sentiment indicator increases the predictive value of a model which uses machine learning mechanism.

The aim of the study was achieved by comparing the quality of the confusion matrix for the model in which the indicator resulting from the sentiment analysis was used with the model without this indicator. Therefore, it must be concluded that sentiment analysis can be really useful in the analysis of stock market data, be it in terms of short-term investments (so-called stock market speculation), or in terms of determining the moment of taking an investment position in a longer period of time. According to the author, his approach can be applied to stock data from various markets, assuming having text data with different people’s entries about the listed companies.