Why So Serious?: Detecting and Tracking Depression Through the Presence of Emoticons
|
If you are the presenter of this abstract (or if you cite this abstract in a talk or on a poster), please show the QR code in your slide or poster (QR code contains this URL). |
Abstract
Background
Depression detection through data mining is no easy task, but measuring the degree of depression is even more difficult to achieve. Many sentiment analysis techniques employ the pragmatic features such as emoticons to detect the sentiment of the text and then to screen depressed people accordingly. However, the ways in which emoticons can be used to predict the mood of the depressed actors is left unexplored.
Objective
Emoticons are textual portrayal of a writer’s mood in the form of icons and serve as a useful tool for online communications of emotions, and the abundance of emoticons found in one’s text can be viewed as a willingness to share his/her feelings. Since emoticons normally carry a humorous and playful tone, writers should refrain from using them when in a serious or depressive mood. Under this assumption, we propose a model predicting the use of emoticons with the amounts of positive or negative emoticons used by the ego's network contacts on two Korean online communities. The first online community is a support group for depressed people, and the other is an information sharing community for people who plan to visit Mexico. In so doing, we show how much significant the emotion diffusion is among depressed people in comparison to people in a non-depressed community.
Methods
We collected data from two separate online communities in Korea. One is an experimental group consisting of people from the depression support community, and the other is a control group consisting of users from the Mexico-related information sharing community. The two groups are similar in size and user activity rate: 192 posts and 1376 replies from 528 users of the depression support community; 383 posts and 3991 replies from 343 users of the information sharing community over 12 month in the year 2012.
We fit random effects time series regression models to predict the use of emoticons in a specific month (t+1) with the amount of emoticons used by the ego's contacts in the previous month (t) in the two different online communities. Our model includes multiple controls such as the ego's network centralities (eigenvector, degree) at t, the number of posts the ego read at t, and the number of post the ego wrote at t.
Results
The time-series random effects model predicting the ego’s use of emoticons over 12 months shows that, in the depression support group, reading friends' posts with positive emoticons increases the emoticon usage of the ego, while reading posts with negative emoticons decreases his/her emoticon usage. In the control group, these two factors turned out to be insignificant. Given that the more frequent use of emoticons is considered as a sign of a brighter mood, only in the depression group, the social contagion of emotions were significant.
Conclusions
In this work, we suggest that the emoticons usage by social contacts can be a good indicator for the ego's depression. Also, collectively using positive emoticons can be used to improve depressive symptoms among the depressed people.
Depression detection through data mining is no easy task, but measuring the degree of depression is even more difficult to achieve. Many sentiment analysis techniques employ the pragmatic features such as emoticons to detect the sentiment of the text and then to screen depressed people accordingly. However, the ways in which emoticons can be used to predict the mood of the depressed actors is left unexplored.
Objective
Emoticons are textual portrayal of a writer’s mood in the form of icons and serve as a useful tool for online communications of emotions, and the abundance of emoticons found in one’s text can be viewed as a willingness to share his/her feelings. Since emoticons normally carry a humorous and playful tone, writers should refrain from using them when in a serious or depressive mood. Under this assumption, we propose a model predicting the use of emoticons with the amounts of positive or negative emoticons used by the ego's network contacts on two Korean online communities. The first online community is a support group for depressed people, and the other is an information sharing community for people who plan to visit Mexico. In so doing, we show how much significant the emotion diffusion is among depressed people in comparison to people in a non-depressed community.
Methods
We collected data from two separate online communities in Korea. One is an experimental group consisting of people from the depression support community, and the other is a control group consisting of users from the Mexico-related information sharing community. The two groups are similar in size and user activity rate: 192 posts and 1376 replies from 528 users of the depression support community; 383 posts and 3991 replies from 343 users of the information sharing community over 12 month in the year 2012.
We fit random effects time series regression models to predict the use of emoticons in a specific month (t+1) with the amount of emoticons used by the ego's contacts in the previous month (t) in the two different online communities. Our model includes multiple controls such as the ego's network centralities (eigenvector, degree) at t, the number of posts the ego read at t, and the number of post the ego wrote at t.
Results
The time-series random effects model predicting the ego’s use of emoticons over 12 months shows that, in the depression support group, reading friends' posts with positive emoticons increases the emoticon usage of the ego, while reading posts with negative emoticons decreases his/her emoticon usage. In the control group, these two factors turned out to be insignificant. Given that the more frequent use of emoticons is considered as a sign of a brighter mood, only in the depression group, the social contagion of emotions were significant.
Conclusions
In this work, we suggest that the emoticons usage by social contacts can be a good indicator for the ego's depression. Also, collectively using positive emoticons can be used to improve depressive symptoms among the depressed people.
Medicine 2.0® is happy to support and promote other conferences and workshops in this area. Contact us to produce, disseminate and promote your conference or workshop under this label and in this event series. In addition, we are always looking for hosts of future World Congresses. Medicine 2.0® is a registered trademark of JMIR Publications Inc., the leading academic ehealth publisher.
This work is licensed under a Creative Commons Attribution 3.0 License.