An analysis of forecasting methods based on sentiment analysis and deep learning for stock market price movements

Predicting the stock market or share market price movement is an important and effective activity. This prediction is a serious challenge as the availability of the data and its types are very volatile. Earlier the movement of the price was subjected to analysis of historical data patterns but recently daily news and sentiments are also important criteria for predicting market trends. The sources of the data are much categorized and need the latest tools and techniques to process it for any analysis. Investors face difficulty in generating profit from the market, leading to the development of prediction theories. This study highlights the review of such mathematical models that include algorithms of Deep Learning and Machine Learning, calculation methods, and performance parameters using historical data and sentiment analysis that are helping in forecasting the price movements of the stock market till date. Accuracy has been the most significant criterion for the prediction algorithms, and neural networks and LSTM have been widely utilised in conjunction with sentiment analysis.