International Journal of Artificial Intelligence and Neural Networks
Author(s) : RADIM BRIXI
The paper presents original results from research that analyses words that are used by the readers of given text to describe the key message in the entire text. Those readers’ words are then lemmatized and compared to lemmatized words from the original main text. The research reveals that for texts with deeper meaning are used surprisingly and also significantly such words by the readers that cannot be found in the entire original text. The percentage of such words increases when forced to express the key message using less words.