Artificial Intelligence Detecting Depression Based on Tweets

Social media platforms can also be a center where research and new technologies are developed in the scientific world, as they also contain great clues about the personalities of those who share. Many of these studies are on Twitter is being done. The latest technology developed by scientists is of great importance.

According to new research published in the journal IEEE Transactions on Affective Computing, scientists whether they are depressed able to detect developed an algorithm. Moreover, the algorithm success rate of almost 90% shared. So how does the algorithm know if a person is depressed?

Users analyzed from 38 data points

According to the shared details, the algorithm can be used on Twitter, such as profile photos of users, content in their posts, sharing times, other users in their social circles. taking a total of 38 data points making an assessment. To be successful, the algorithm was trained with two different databases that contain the Twitter history of thousands of Twitter users and display their mental health status.

While 80% of the data in the database was used to train the algorithm, the remaining data was kept for testing the algorithm. As a result of the tests made with this data, which the algorithm has never seen before, the algorithm has 89% success announced. In John Hopkins University’s CLPsych 2015 data, this success rate is to 71% declined.

In the event that the algorithm in question is developed, scientists are aware of the fact that social media such as Twitter and Facebook that you can constantly monitor your mental health for preventive purposes. explained. On the other hand, although the data used in this study is given at the request of the user, the fact that not everyone wants to share their data in the future has already started the data discussions.