Currently there is only one mood tag "angry", which is handled with keywords and some simple rules. With more data, more complex and interesting rules can be de

Currently there is only one mood tag "angry", which is handled with keywords and some simple rules. With more data, more complex and interesting rules can be derived. Then the rage quotient is calculated, i.e. what percentage of people are angry (actually, it takes the number of angry posts in the sample and divides by half the sample size). The quotient is then mapped unto a list of verbal descriptions. This means that I can add or remove description as needed.