Description
Abstract: This study analyses Russian-language comments posted on YouTube videos about the war in Ukraine by using text mining methods. The analysis results suggest the possibility to influence what individuals think and believe during the war by changing the context in digital public space.
Data and Methods: The analysis subject is Russian comments posted on the 10 most-viewed YouTube videos collected by searching videos for “украина (Ukraine)” and “война (war)” and filtering the results by highest view count.
Results (Co-occurrence Analysis): Figure 1 “Human” and “Life” related words are displayed such as 'human', 'family', 'Christ' or 'time'. At the same frequency “Country and Place Name” related words are displayed such as 'USA', 'Kiev', 'West' or 'Poland'.
Results (LDA analysis): Figure 2 shows “Military and War” related words are given low polarity scores and classified as negative such as 'strike/attack', 'submarine' and 'ballistic missile' in comments regarding USA. In comments regarding Germany, Figure 3 shows “Nazi” related words are given low polarity scores and classified as negative such as 'nazi', 'Stepan Bandera' and 'Hitler'. In comments regarding China, Figure 4 shows “Economy and bilateral” related words are given high polarity scores and classified as positive such as 'corporation', 'locomotive' and 'intimate'.
Conclusion: It turns out Russian users tend to associate themselves with "connections between people" or "faith in God" when referring to the war in Ukraine, but the users display a more aggressive attitude toward USA by mentioning specific names of strategic missiles or military weapons. The analysis results suggest the possibility to influence what individuals think and believe during the war by changing the context in digital public space.
