A set of methods to analyze social media content (Twitter, Youtube, VKontakte, Yandex.news) which I used in my PhD work. The methods are isolated as separate R files which are not yet cleaned to be uploaded. Some of them require user/password to be entered for social networks. Some of the require some form of manual mods before results can be obtained.
Here is example of analytics of youtube channels. We can compare on xray chart how often guests of show talk about Chernyshevsky and Dostoevsky:
Collecting all the subtitles allows us to make word collocation plots like the one below:
For twitter and VKontakte data we can compare group of fiction writers and see in which context and sentiment people tend to speak about them:
