Looking for Signs of Depression on Twitter with Natural Language Processing


Elizabeth Harrington of Free Beacon reports, "The federal government is studying how to use Twitter for surveillance on depressed people. The University of California, San Diego (UCSD) began a study financed by the National Institutes of Health last month that will provide 'population level depression monitoring' through the social media site. The project, 'Utilizing Social Media as a Resource for Mental Health Surveillance,' is costing taxpayers $82,800."

Harrington continues, "While Twitter has been used by government agencies, such as the Secret Service and Department of Homeland Security, for national security related monitoring, the project suggests the social network can be used for public health surveillance as well. 'Major depressive disorder is one of the most common debilitating illnesses in the United States, with a lifetime prevalence of 16.2 [percent],' the project grant states… 'We propose using twitter and [Natural Language Processing] NLP as a cost-effective and flexible approach to augmenting current telephone- based surveillance methods for population level depression monitoring,' the grant said."

He adds, "The researchers will create algorithms to determine if people are depressed through their tweets, which they hope will serve as a basis for monitoring mental illness. They will also engage with depressed individuals on Twitter directly. 'Developing these algorithms and resources will provide the bedrock for building social media based surveillance systems,' the grant said."

Read more here.

Image: Courtesy Flickr