A corpus of Ukrainian and Russian news was found and each text was annotated with three categories: positive, negative and neutral. Various techniques used for Sentiment Analysis were compared and the highest accuracy score was for the Support Vector Machine Model.