April 7, 2022

Bot to Spot Depressed Twitter Users

April 7, 2022

Bot to Spot Depressed Twitter Users

April 7, 2022

Researchers at Brunel University in London have developed an algorithm which can spot depression in Twitter users with almost 90% accuracy.

They claim to be able to determine a users state of mind by analyzing 38 data points from a public Twitter profile such as post content, post times and even other users in the social circle pertaining to that profile.

The research team claim that their algorithm has many practical applications; including early diagnosis of depression, police investigations and employment screening.

Professor Abdul Sadka, Director of Brunel’s Institute of Digital Futures stated that the algorithm was tested against other depression detection techniques and “In all cases, we’ve managed to outperform existing techniques in terms of their classification accuracy.”

Professor Sadka went on to state “Anything that’s above 90% is considered excellent in machine learning. So, 88% for one of the two databases is fantastic.”

The researchers also claim that such a system could be used by social media platforms such as Twitter and Facebook to flag mental health concerns with the affected users themselves.

As a consequence, mental health crises might be avoided when a user is given the chance to react to warning signs that they may not have noticed; perhaps prompted to visit their doctor or remminded to take their medication.

Prof Huiyu Zhou, Professor of Machine Learning at the University of Leicester, stated “The proposed algorithm is platform independent, so can also be easily extended to other social media systems such as Facebook or WhatsApp.”

© Vanilla Websites

Blog category: News & Events

Bringing you the very best in bespoke, business-class website design at affordable prices. Your perfect website is our job.


(0151) 528 2181