With its millions of users, Twitter has grown into a huge channel of personal expression, which can be useful in public health issues – especially depression.
Microsoft Research Redmond, Eric Horvich, is a pioneer of Twitter and depression research, and estimates that at some point smart systems will be able to analyze a user's Twitter feed and warn him if he is at risk of depression, according to a report of Time.
"We wondered if we could create tools that would be able to tell when someone is depressed, just by the publications του. Τι λένε οι άνθρωποι στον κόσμο σε δημόσιους χώρους; Μπορείτε να φανταστείτε tools που θα ενημερώνουν για μεταβολές στη διάθεση ενός χρήστη, πριν καν τις νιώσει ο ίδιος» αναφέρει.
Horvich and a team of researchers have helped develop a method that allows depictions to be depicted on Twitter users with 70% accuracy by scraping their tweets. However, the method still has much room for improvement, as some clues can be avoided and people who may be experiencing this problem may not be identified. Also, according to Hornic, there is a mistaking issue, as 10% of cases found that depressed healthy users were at risk.
As part of the research, the Microsoft team found 476 Twitter users, 171 of whom were depressed. The researchers then looked at their Twitter history up to one year before diagnosis, looking for various signs of depression, analyzing 2,2 million tweets with computational models. By comparing the tweets of users depressed with these healthy users, a method has been developed that can predict depression before they occur. This "model" was then tested in another sample of users with a success rate of 70%.
Some tweets were "obvious": "I want someone to hug me and be there for me when I'm sad," "having a job again makes me happy. Less time to be sad and to see sad movies, "etc. But Microsoft researchers also looked at other factors such as the number of tweets per day, while users made tweets, the degree of interaction with others , the type of language, etc.
They also searched for words – wrenches indicating depression ("anxiety", "nausea", "sleep", "nervousness", etc.) were often used, but there were other, seemingly more "innocent" words, such as "love", "he", "she" , "home", "tolerance", "songs", "movies" etc.). The frequency of tweets and conversations with other users is also important, as depressed people tend to tweet less and interact less with other people, according to Horwitz.
One area in which such research might be useful is to evaluate public reactions to major events. Observing and analyzing Twitter feeds after "traumatic" incidents could allow a further understanding of exactly how users are affected by news.
"Our view is that Twitter is the largest observational study of human behavior we have ever seen, and we are working hard to exploit it" says Time Tyler McCormick of the Center for Statistics and Social Sciences of the University of Washington. It is noted that McCormick's team is also working on the subject, as well as a team at the University of California-San Diego.
