ALGORITHM FOR CALCULATING THE RATING OF A POST PLACED IN VIRTUAL COMMUNITIES

Abstract

The increase in the number of virtual communities and their intensive development causes an increase in the number of users and the amount of data due to communication between them. The data posted in the communities is based solely on the experience of the authors who published them. As a result, there is a need to verify the posted data for validity. One of the ways to check the validity of posted information is to calculate the rating of the post. The rating of a post should be determined taking into account all possible reactions that users can leave: evaluations, interactions and comments. Each of the given reactions has a certain set of criteria: evaluation (like, dislike, love, care, haha, wow, sad, angry, in the rate), interaction (share, save and views) and comment on the post. After conducting an analysis more than twenty virtual communities, it was found that most of them don’t have the entire set of criteria, which affects the calculation of reactions. Finding certain regularities, three scenarios were built for calculating the value of evaluation and interaction reactions. To determine the reaction to commenting, sentiment analysis of the comments to the post was carried out. This analysis is intended for automated detection of emotionally colored vocabulary in texts and emotional assessment of authors in relation to the informational content of posts. This sentiment analysis provides a quantitative assessment of positive, negative and neutral emotions. Neutral emotions were not taken into account. As a result, an algorithm was built that shows the sequence of actions for calculating the rating of a post. The developed algorithm is universal for any community because all reactions in virtual communities are taken into account to calculate the rating of a post.