How are you feeling today? Don’t worry, the internet will soon be able to answer that question for you.
Researchers at the CSIRO today unveiled software that will mine real-time data from Tweets to create “emotional maps” of cities and even countries around the English-speaking world. The software can analyse around 500 Tweets a second, and up to 27 million over 24 hours, which roughly corresponds to the number of Tweets your average bored office worker sends out on a daily basis.
The software, designed as a part of the “We Feel” project, analyses the emotional tone of Tweets, taking into account 600 words that correspond to specific emotional states, from joy to sadness, surprise to fear and love. It will attempt to chart how emotions in different regions of the world fluctuate over specific time periods, and will likely improve current mental health research methods.
For example, when this screengrab was taken this afternoon, the preceding 12 hour period had seen over 3.6 million Tweets related to ‘sadness’, or roughly 3.15% of all Tweets in English.
Hopefully future versions of the program will be able to hone in in more complex emotional states, like guilty joy upon discovering a packet of uneaten Peanut Butter M&Ms in the back of the cupboard, determination to go back to the gym this week and uncertainty as to whether or not it’s socially acceptable to tell people you find Ramsay Snow kind of attractive.
