Sir, i need help, how can i determine the number of tweets in the dataset that can be classified as happy
Hi, you have to identify all possible synonyms that we humans use to express happiness on social media, such as “great!” , “yay!”, and so on…
Basically you will need a bag of such words , so any text in the tweet matches to one of the words from your happiness bag of words , then classify that tweet as happy.
This is a very basic way to start. Over this you can build up, by using number of positive expressions and negatives expressions such as ‘not’ ‘bad’ and so on. to refine your classification method to truly understand in human context if the tweet means happiness. This is just an idea, you can experiment with.