Abstract
Recently, people have come to commit suicide due to excessive stress received in work or bullying. Since some suicide victims were suffering from depression, it is thought that there is a close relationship between suicide and depression. As far as emotions before suicides, you can see them from posts on SNSs, such as Twitter. Therefore, in this study, we collected tweets of Twitter users who suffer from mental disorder and normal individuals. Then we calculated emotional values of the tweets by setting numerical values of emotions to each word, and compare these values statistically. As a result, the emotional values showed different tendencies it suggests that this method is effective for estimating depression.
