Semi-supervised User Geolocation via Graph Convolutional Networks

04/22/2018

∙

by   Afshin Rahimi, et al.

∙

0

∙

share

Social media user geolocation is vital to many applications such as event detection. In this paper, we propose GCN, a multiview geolocation model based on Graph Con- volutional Networks, that uses both text and network context. We compare GCN to the state-of-the-art, and to two base- lines we propose, and show that our model achieves or is competitive with the state- of-the-art over three benchmark geoloca- tion datasets when sufficient supervision is available. We also evaluate GCN under a minimal supervision scenario, and show it outperforms baselines. We find that high- way network gates are essential for control- ling the amount of useful neighbourhood expansion in GCN.