ADVERTISEMENT
Artificial Intelligence Analyzes Online Posts for Depressive Symptoms in Young People
Artificial intelligence technology trained to analyze content in online posts written by young people demonstrated a 98% to 99% mean performance in detecting depression symptoms, according to a study published in the journal Neural Computing and Applications.
“The researchers have applied artificial intelligence algorithms to recognize and make comparisons between words, sentences, and expressions, and the context that links them together,” lead study author Zia Uddin of the research group SINTEF Digital in Norway told Norwegian SciTech News. “The system is now able to recognize signs of depression and melancholy, and this will open opportunities for the development of a digital solution that can provide urgently needed help to sufferers.”
The algorithm focused on posts to Ung.no, a popular Norwegian website where young people can ask questions and access information about school, relationships, and other issues. The content is quality assured by experts in relevant fields.
Related: Air Pollution Associated With Increased Depressive Symptoms in Adolescents
“When they ask questions on the website Ung.no, they often tell the story of their previous life experiences,” said study coauthor Kim Kristoffer Dysthe, a mental health provider and medical doctor, in the Norwegian SciTech News report. “They then go on to describe the events that are triggering the thoughts they have about their current situation and end up by talking about the behaviors and symptoms these events have caused. Often, this content is easily recognizable as a sign of depression.”
Researchers trained the algorithm to identify terms, sentences, and questions that may signal depressive symptoms. Compared with conventional artificial intelligence approaches based on general word frequencies (“some topmost frequent words are chosen as features from the whole-text dataset and applied to model the underlying events in any text message,” researchers explained in the study), this approach focused on specific depressive symptoms.
When researchers evaluated their artificial intelligence approach, it achieved a mean prediction performance of 98% on the first dataset of 11,807 texts and 99% on the second dataset of 21,807 texts compared with 91% mean recognition performance achieved by conventional approaches, according to the study.
“Although the proposed approach is applied on a Norwegian dataset, a similar robust approach can be applied on other depression datasets developed in other languages with proper annotations and symptom-based feature extraction,” researchers wrote. “Thus, the depression prediction approach can be adopted to contribute to develop better mental health care technologies such as intelligent chatbots.”
References
