A new software program can tell whether a blogger is depressed.
The program scans online text and blogs, identifying not only obvious keywords, but even subtle clues that can determine the blogger’s psychological state.
“The software program was designed to find depressive content hidden in language that did not mention the obvious terms like depression or suicide,” said Professor Yair Neuman of Ben Gurion University of the Negev, Israel, who led the study.
“A psychologist knows how to spot various emotional states through intuition. Here, we have a program that does this methodically through the innovative use of ‘web intelligence,’” said Neuman.
Many people with depression are undiagnosed. While there are a number of screening tools available on the Internet, only individuals who suspect they may have depression are likely to use these tools.
To design the program, named Pedesis, Professor Neuman and his team first used extensive English language web searches to define word patterns associated with depression, analyzing the text associated with the search term: “depression is like….”
For example, the program might detect words that express different emotions, like colors that the writer uses to metaphorically describe a situation, such as “black.” Other concepts associated with depression that might trigger the software include terms like sleep deprivation, or loneliness.
To test the program they analyzed more than 350,000 texts from 17,031 bloggers. (Permission from the bloggers was obtained prior to the study.) The Pedesis software identified the 100 most depressed and the 100 least depressed bloggers.
A panel of four clinical psychiatrists reviewed the posts from each group and found that there was a 78 percent correlation between their clinical impressions based on the texts and the software’s diagnosis.
Computer-generated diagnosis or online therapy is not a substitute for human interaction, but may provide a useful screening tool. “No one can actually replace excellent human judgment,” says Neuman. “The problem is that most people are not aware of their situation and they will never get to an expert psychologist.”
Although the Pedesis program was originally designed for academic purposes, Neuman hopes that these findings could be useful clinically, to screen for potential suicides. If implemented on a widespread basis, a screening process might increase the user’s awareness of his or her condition, and provide recommendations. If the blogger agrees, he or she could seek professional help.
Professor Neuman’s results will be presented at the 2010 IEEE/WIC/ACM International Conference on Web Intelligence and Intelligent Agency Technology in Toronto.
Source: Ben Gurion University