Abstract
Sentiment analysis is a tool used in many areas (marketing, management,
sociology) and its effectiveness depends on the quality of sources and
the volume of data gathered. The authors present procedures that allow
the collection of large volumes of text data and the processing thereof,
the result of which is a quick (practically in real time) assessment of
emotions and their changes. In the example analysis Russian sources are
given (in the Russian language, creating a system of media monopolised
by the authorities). The analysis was between week 1 and 17 of 2022 -
before and during the conflict between Russia and Ukraine. The results
obtained, trends of changes in sentiments pertain to Presidents Joe
Biden, Andrzej Duda, Aleksandr Lukashenko, Emmanuel Macron, Volodymyr
Zelensky and Vladimir Putin - the results are emotionally charged
figures of presidents created for the internal purposes of Russia and
provide significant information for analysts of the ongoing conflict.
Obtaining similar results without the use of information technology is
practically impossible. Sentiment analysis of large resources makes it
possible to detect the change dynamics that are difficult for a human
being to capture (a person analysing source materials). Concurrently,
the cost and time of obtaining results recommends the described method
as supporting analytical work.