ALGORITHM FOR AUTOMATED DETERMINATION EMOTIONAL COLORING OF TEXT DATA

Abstract

This article is devoted to solving the problem of the automated computation of emotional coloring of natural language text. In this case a combined approach is used. This approach provides the categorization of text data and the analysis of syntactic relations between words in this text. It enables to consider the emotional color (tonality) of individual words, but the key groups of syntactically related words. These groups are composed of emotionally colored words and words that amplify or weaken the emotional coloring of other words within the group. This approach provides the ranking of text data into categories of emotions and enables to add new categories of emotions. An algorithm of the automated computation of emotional coloring of natural language text based on this approach is developed.