The Facebook Timeline reveals far more than the destination, the relationship status or the Job. According to a recent study, the contributions could deliver on the platform, with information as to whether a person becomes ill in the following months of Depression. Therefore, the choice of words for depressed people is often different from that of other people.
As the basis of the data of 683 patients who had come due to an illness in the emergency Department of the University of Pennsylvania were used by the researchers. All participants agreed that the researchers were evaluating in addition to your medical record and your Facebook Postings. In total, the researchers analyzed more than 500,000 posts on the Social Media platform, and brought them in connection with the medical history.
I More perspective, more physical complaints
114 of 683 participants had a diagnosed Depression, as the researchers in the journal “PNAS” reports. Your word choice was different in several respects from that of other patients:
people with depression used in the analysis according to the first person perspective propagated . Specifically, the words “I” and “me before came in your Postings more often”. The talk, that the employees Concerned much with themselves, the researchers write. In your Postings you described also increasingly the characteristics of a depressed mood , it was more likely to solitude . Specifically, the made by expressions such as “tears”, “crying” and “pain”, but also “miss”, “lost” and “alone”. To the physical complaints were in people with depression more of an issue. Concretely, this is reflected in terms such as “pain”, “head”, “hospital”, “surgery”, “tired” and “sick”. This is explained by the fact that depression often go hand in hand with physical Suffering, the scientists write. In addition, the depressed participants were also more likely to hostile , what the researchers made of terms such as “hatred,” “Shame” and “smh” (abbreviation for “shaking my head”, so a shake of the head).
The results are consistent with those of previous studies. Several other studies also came to the conclusion that people with depression pronouns more frequently, the personal “I” and “me”. Most of the previous analyses were, however, focused on private texts and not on the clear public messages more in social media.
algorithm realized 70 percent of the depression
Using the data, the researchers programmed a algorithm that should detect depression solely on the basis of the Facebook Timeline. As the Basis of the contributions from the six-month period served prior to the diagnosis. Although the Computer identified the depression, only 70 percent of the people Affected. The hit rate was, however, similarly high as in conventional surveys, the researchers write.
the scientists Evaluated, however, the Postings, the data were written more than half a year before the diagnosis, recognized by the Computer, the depression. This result suggests that the disease has changed the word choice of the parties Concerned in the course of time, actually.
The researchers hope their work eventually in fact, people with depression can help. Although the disease is very common and easy to treat, it remains often undetected, especially in men. Conceivable is, for example, that the evaluation of the social media represents a first stage in the in-house doctor care, the scientists write in their study. In the case of a suspicious result, the physician could follow up then.
MORE on THE TOPIC of depression, men suffer differently
“people write in social media about aspects of your life, to which we would in the medicine and in the research that would otherwise be difficult to access,” writes Andrew Schwartz of the University of Pittsburgh. This opens up possibilities power privacy but also all the more important.
“such An algorithm could write Social Media posts in to protect health and transform information,” the authors in the conclusion of your study. “With this challenge need to contact the developer and decision-makers apart.” The authors call for stringent guidelines to regulate the access to the data and to leave the decision on own health care always in the patient.
