In a new study, researchers have revealed that your tweets can tell whether you are happy or not. A team of computer analysts including an Indian-origin has developed an algorithm that can accurately predict the level of satisfaction and happiness by just accessing the twitter account of an individual.
Scientists from the Unversity of Iowa have developed an algorithm that uses word count of a tweet to measure the life satisfaction level and study authors believe that a tweet can reflect person’s state of mind.
Unlike previous social media studies that tell how a person feels at the particular moment, the current study tells how users feel about their life.
For the study, Chao Yang, a graduate at the University of Iowa (UI) in US along with Indian-origin Padmini Srinivasan, a professor at UI, analysed 3 billion tweets from October 2012 to October 2014. Study authors selected only those tweets that consisted words like “I”, “me” and “mine” in order to get a better self-reflection of person’s state of mind.
Researchers then developed an algorithm that checked the level of satisfaction or dissatisfaction in one’s life. It was uncovered that tweets of satisfied and happy people remained steady and stable over a long period of time and it remained unaffected by events like famines, earthquakes, any major sports tournament, elections, etc.
On the contrary, previous studies that took short-term happiness into account had found that external events affect people’s daily moods heavily.
In the study, it was found that unhappy people frequently used personal pronouns, conjunctions and profanity in their tweets. Dissatisfied people are 10 percent more likely to use negative expression in their tweets and use words like “should,” “would,” “expect,” “hope,” and “need” that may express determination and aspirations for the future.
While satisfied and happy people frequently use more than 140 characters in their tweets and are likely to use positive words and talk on positive topics like health, sexuality, money and religion.
The study appeared in the journal PLOS One.