The insights gleaned from Facebook posts can tell if people are suffering from diabetes and other diseases, according to an American study.
The researchers found that about 1,000 people surveyed the language used on the platform, compared to social psychological information, could have a better key word in identifying a particular health condition.
Diabetes is one of the predictable conditions. The researchers did not specify whether they were type 1 diabetes or type 2 diabetes, as well as anxiety, mental illness and depression.
Researchers at Penn Medicine and Stony Brook University believe that the findings could lead to new applications using medical artificial intelligence.
If the language used on the social media platform gets the consent, it can be monitored sometime like the physical state.
"Although the study ended prematurely, the insights gleaned from this post can be used to provide information to patients and providers about their health," said Dr. Raina Merchant, director of Penn Health's Digital Health Center. Research.
"This information can provide additional information on disease management and aggravation, because social media posts are often information about someone's lifestyle choices and experiences or the way they feel."
Special automated data collection techniques are used to allow teams to review all Facebook posts for each participant whose health history is also analyzed.
A total of 21 conditions were investigated, and the results indicate that all conditions were predicted by Facebook activity. The survey found that Facebook data was more effective in predicting 10 situations than demographic information.
According to the results, some of the languages were "drink" and "bottle" linked to alcoholism but apparently other words were not so clearly associated with health status . People who use the words "God" and "prayer" are 15 times more likely to get diabetes than users who use the term the least.
The study was published in the PLOS ONE journal.