Forecasting the direction of stocks is a crucial work for all investors seeking profits. Stock forecasting is critical because precisely anticipating stock prices can lead to desirable rewards for investors who make sound decisions about purchasing and selling stocks at the right moment. There are three types of datasets that are often used for stock market prediction: financial (historical) data, sentiments (social media data and financial news), and the combination of financial data and sentiments. The suggested work used the recurrent neural network (RNN) algorithm to anticipate stock prices by combining financial and sentiment data for the next day's close. The experiment is carried out on stocks that are traded globally. When compared to other models, the proposed model outperformed in terms of low error rate and high accuracy.