DocumentCode :
3740135
Title :
Predicting Depression of Social Media User on Different Observation Windows
Author :
Quan Hu;Ang Li;Fei Heng;Jianpeng Li;Tingshao Zhu
Volume :
1
fYear :
2015
Firstpage :
361
Lastpage :
364
Abstract :
Depression has become a public health concern around the world. Traditional methods for detecting depression rely on self-report techniques, which suffer from inefficient data collection and processing. This paper built both classification and regression models based on linguistic and behavioral features acquired from 10,102 social media users, and compared classification and prediction accuracy respectively among models built on different observation windows. Results showed that users´ depression can be predicted via social media. The best result appears when we make prediction in advance for half a month with a 2-month length of observation time.
Keywords :
"Feature extraction","Media","Pragmatics","Correlation","Psychology","Predictive models","Training"
Publisher :
ieee
Conference_Titel :
Web Intelligence and Intelligent Agent Technology (WI-IAT), 2015 IEEE / WIC / ACM International Conference on
Type :
conf
DOI :
10.1109/WI-IAT.2015.166
Filename :
7396831
Link To Document :
بازگشت