ОЦІНКА ЕФЕКТИВНОСТІ МЕТОДІВ АНАЛІЗУ ЕМОЦІЙНОГО ЗАБАРВЛЕННЯ КОМЕНТАРІВ

This thesis is devoted to the study and analysis of various methods of identifying and classifying the emotional coloring of comments in modern Internet resources. The research is aimed at revealing effective approaches to automatic detection of emotional responses in textual content, taking into account the specificity of detecting different shades of emotions, such as positive, negative and neutral. The analysis and comparison of different methodologies and algorithms will allow to determine the optimal approaches for the development of tools for the automated analysis of large volumes of textual information, taking into account its emotional component.