A Unified Neural Network Model for Geolocating Twitter Users

Proceedings of the 22nd Conference on Computational Natural Language Learning (CoNLL 2018), pages 42–53 Brussels, Belgium, October 31 - November 1, 2018. c⃝2018 Association for Computational Linguistics 42 A Unified Neural Network Model for Geolocating Twitter Users Mohammad Ebrahimi, Elaheh ShafieiBavani, Raymond Wong, Fang Chen University of New South Wales, Sydney, Australia Data61 CSIRO, Sydney, Australia {mohammade,elahehs,wong,fang}@cse.unsw.edu.au Abstract Locations of social media users are impor- tant to many applications such as rapid disas- ter response, targeted advertisement, and news recommendation. However, many users do not share their exact geographical coordinates due to reasons such as privacy concerns. The lack of explicit location information has mo- tivated a growing body of research in recent years looking at different automatic ways of determining the user’s primary location. In this paper, we propose a unified user geoloca- tion method which relies on a fusion of neural networks. Our joint model incorporates dif- ferent types of available information includ- ing tweet text, user network, and metadata to predict users’ locations. Moreover, we uti- lize a bidirectional LSTM network augmented with an attention mechanism to identify the most location indicative words in textual con- tent of tweets. The experiments demonstrate that our approach achieves state-of-the-art per- formance over two Twitter benchmark geolo- cation datasets. We also conduct an abla- tion study to evaluate the contribution of each type of information in user geolocation perfor- mance. 1 Introduction Knowing physical locations involved in social me- dia data helps us to understand what is happening in real life, to bridge the online and offline worlds, and to develop applications for supporting real-life demands. For example, we can monitor public health of residents (Cheng et al., 2010), recom- mend local events (Yuan et al., 2013) or attractive places (Noulas et al., 2012) to tourists, identify locations of emergency (Ao et al., 2014) or even disasters (Lingad et al., 2013), and summarize re- gional topics (Rakesh et al., 2013). Even though platforms such as Twitter allow users to geolocate their posts to reveal their locations either manually or with the help of GPS, it is reported that less than 1% of Twitter data has geo-coordinates provided (Jurgens, 2013). Moreover, location information on Twitter is far from being complete and accu- rate. For instance, self-declared home information in many user profiles is inaccurate or even invalid (Hecht et al., 2011). The lack of explicit location information in the majority of tweets has moti- vated a growing body of research in recent years looking at different automatic ways of determining the user’s primary location (i.e.,user geolocation) and/or - as a proxy for the former - the location from which tweets have been posted (Ajao et al., 2015). Geolocation methods usually train a model on a small set of users whose locations are known (e.g., through GPS-based geotagging), and pre- dict locations of other users using the resulting model. These models broadly fall into three cat- egories: text-based (Eisenstein et al., 2010; Wing and Baldridge, 2011; Roller et al., 2012), network- based …

Content truncated to 500 words