Moodpanda.com is a mood and Sentiment tracking site so when the user posts "I Love #coke" or "#coke is sick" it is attached to there chosen mood number (1-10) so they can get a exact rating of how the person feels about the product or brand etc. It is direct Sentiment analysis and does not fall foul of the normal issues with Sentiment analysis e.g. is the word "sick" good or bad? Different ways of phrasing etc. are not an issue for moodpanda.com
I'd like to know how they measure the mood. Sentiment analysis is a bit of a black box.