The concept is quite interesting, but how is that possible to find out which users' messages are real marks for definite piece of news to be truth or lie, and which users' messages are just the same prepaid fake point of view?

Do you need any help with russian or english audience? As I've mentioned, the idea itself is great.