| Resumen |
Deriving intelligence from text is important as it can provide valuable information on how events influence public opinion. In this work, a classification task was done in order to obtain the sentiment behind the polarity of an economic text using machine learning and deep learning methods. We analyzed the text for keywords that can be categorized into positive, negative and neutral reviews and found more insights. In the final result of classifying three groups (positive, negative and neutral), the models were unable to perform up to 80% accuracy, where only one variant has the accuracy of 80% as the best on the test dataset. © 2020, Springer Nature Switzerland AG. |
