Unifying Text, Metadata, and User Network Representations with a Neural Network for Geolocation Prediction

Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics, pages 1260–1272 Vancouver, Canada, July 30 - August 4, 2017. c⃝2017 Association for Computational Linguistics https://doi.org/10.18653/v1/P17-1116 Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics, pages 1260–1272 Vancouver, Canada, July 30 - August 4, 2017. c⃝2017 Association for Computational Linguistics https://doi.org/10.18653/v1/P17-1116 Unifying Text, Metadata, and User Network Representations with a Neural Network for Geolocation Prediction Yasuhide Miura†,‡ yasuhide.miura@fujixerox.co.jp Motoki Taniguchi† motoki.taniguchi@fujixerox.co.jp Tomoki Taniguchi† taniguchi.tomoki@fujixerox.co.jp Tomoko Ohkuma† ohkuma.tomoko@fujixerox.co.jp †Fuji Xerox Co., Ltd. ‡Tokyo Institute of Technology Abstract We propose a novel geolocation prediction model using a complex neural network. Our model unifies text, metadata, and user network representations with an attention mechanism to overcome previous ensem- ble approaches. In an evaluation using two open datasets, the proposed model exhib- ited a maximum 3.8% increase in accuracy and a maximum of 6.6% increase in ac- curacy@161 against previous models. We further analyzed several intermediate lay- ers of our model, which revealed that their states capture some statistical characteris- tics of the datasets. 1 Introduction Social media sites have become a popular source of information to analyze current opinions of nu- merous people. Many researchers have worked to realize various automated analytical methods for social media because manual analysis of such vast amounts of data is difficult. Geolocation prediction is one such analytical method that has been studied widely to predict a user location or a document location. Location information is crucially important information for analyses such as disaster analysis (Sakaki et al., 2010), dis- ease analysis (Culotta, 2010), and political anal- ysis (Tumasjan et al., 2010). Such information is also useful for analyses such as sentiment analysis (Mart´ınez-C´amara et al., 2014) and user attribute analysis (Rao et al., 2010) to undertake detailed region-specific analyses. Geolocation prediction has been per- formed for Wikipedia (Overell, 2009), Flickr (Serdyukov et al., 2009; Crandall et al., 2009), Facebook (Backstrom et al., 2010), and Twitter (Cheng et al., 2010; Eisenstein et al., 2010). Among these sources, Twitter is often preferred because of its characteristics, which are suited for geolocation prediction. First, some tweets include geotags, which are useful as ground truth locations. Secondly, tweets include metadata such as timezones and self-declared locations that can facilitate geolocation prediction. Thirdly, a user network is obtainable by consideration of the interaction between two users as a network link. Herein, we propose a neural network model to tackle geolocation prediction in Twitter. Past studies have combined text, metadata, and user network information with ensemble approaches (Han et al., 2013, 2014; Rahimi et al., 2015a; Jayasinghe et al., 2016) to achieve state-of-the-art performance. Our model combines text, metadata, and user network information using a complex neural network. Neural networks have recently shown effectiveness to capture complex represen- tations combining simpler representations from large-scale datasets (Goodfellow et al., 2016). We intend to obtain unified text, metadata, and user network representations with an attention mecha- nism (Bahdanau et al., 2014) that is superior to the earlier ensemble approaches. The …

Content truncated to 500 words