Nowadays—especially since the pandemic started—mental health has become a struggle for a majority of people, particularly stress, anxiety, and depression.
Fortunately, scientists have now developed a recent algorithm on Twitter that has up to 90% accuracy in detecting depressed individuals. The journal, IEEE Transactions on Affective Computing, explains the algorithm works by analyzing 38 data points from a user’s public profile.
The analysis includes the user’s posts, the period in time they made these posts, as well as other users that interact with them. This new algorithm has been tried using 2 databases containing the Twitter history of thousands of users along with their mental health status.
They then use 80% of this information to teach the bot the detection process. The bot filters out users with less than 5 tweets and also corrects for misspellings. In the end, the bot considers 38 factors such as positive and negative vocabularies used along with the kind of emojis that accompany them.
- Oxford study says there’s no increased risk of brain tumor for cellphone users
- World’s fastest AI supercomputer is building a 3D map of the universe
- Scientists discover protein that can alter, forget memories
Scientists also claim that the system may flag a user’s depression a day before he posts something in public. This may also start a way for fellow social media platforms like Facebook to proactively flag users with growing mental health concerns.
While it may not be totally 100% yet, scientists are confident that the closer they hit a 90% figure already shows an improvement; this in turn could further grow in the future as technology continues to expand.
With a lot of people’s mental health going undetected, it is concerning that anxiety and depression might worsen in the long run. Thanks to the kind of technology being developed today, we may have the opportunity to save our family and friends from their mental health struggles.
Via: The Independent