Investigation of the Online Press and Commodity Exchange Using Neural Networks

Abstract

The printed press greatly influences people's consumption attitudes, their stock market decisions, and their information on the decisions and internal problems of certain organizations and companies. In other words, the information obtained in this way may encourage them to give up a certain stock market position or to establish new ones, as they think they are familiar with the willingness and tendencies to buy on the market. Therefore, it is important to analyze how close this relationship is, and how effectively the content of articles published in the press can be linked to the prediction of short-term exchange rate or price developments, and to see whether a formal method can be applied to investment decisions. With the help of neural networks, we can search for relationships between large amounts of diversified data sets, and to set up a contingent forward-looking price and/or exchange rate forecasting model.