To understand the mood within a text, whether the text is dark and gloomy, or just happy, we need to read it. But what if we want to find out whether songs written in the 90’s were happier than songs these days?
Instead of spending days on reading lyrics from hundreds of songs, we can let the computer figure it out for us. To do so, we give each word a value and then use simple statistics to determine the mood.
But how do we know the value of a word? Fortunately, there are researchers who work with finding these values, where they ask a bunch of people to rate words. Two good collections of such word lists are ANEW and AFINN.
What countries write about each other
In a project where I applied ANEW, we were looking at what different countries were writing about each other on twitter. We downloaded tweets from areas in each country, converted all words to their corresponding values, and then displayed them on a map.