Economic forecasting database using GDELT project data

In this thesis, GDELT Global Knowledge Graph (GKG) data is used to analyse the sentiment values of our economic dictionary and for individual countries in a specific timeframe. The tone values of the individual online articles were used as basis for sentiment values for individual economic terms, which are then used on the original file resources to find their total sentiment values, negative sentiment values, positive sentiment values, and polarity of the files. Polarity of the files are calculated using the amount of matching terms divided by total terms in files. First 5 sample results of all these fields are calculated for individual files and location tags in the files initially by using the terms individually, and then by clustering the terms in groups.