Spatial Statistics of Term Co-occurrences for Location Prediction of Tweets

Predicting the locations of non-geotagged tweets is an active research area in geographical information retrieval. In this work, we propose a method to detect term co-occurrences in tweets that exhibit spatial clustering or dispersion tendency with significant deviation from the underlying single-term patterns, and use these co-occurrences to extend the feature space in probabilistic language models. We observe that using term pairs that spatially attract or repel each other yields significant increase in the accuracy of predicted locations. The method we propose relies purely on statistical approaches and spatial point patterns without using external data sources or gazetteers. Evaluations conducted on a large set of multilingual tweets indicate higher accuracy than the existing state-of-the-art methods.

[1]  Lars Backstrom,et al.  Find me if you can: improving geographical prediction with social and spatial proximity , 2010, WWW '10.

[2]  Brendan T. O'Connor,et al.  A Latent Variable Model for Geographic Lexical Variation , 2010, EMNLP.

[3]  Steven Schockaert,et al.  Spatially Aware Term Selection for Geotagging , 2014, IEEE Transactions on Knowledge and Data Engineering.

[4]  Kyumin Lee,et al.  A content-driven framework for geolocating microblog users , 2013, TIST.

[5]  Bruno Martins,et al.  Automated Geocoding of Textual Documents: A Survey of Current Approaches , 2017, Trans. GIS.

[6]  Kyumin Lee,et al.  Uncovering social spammers: social honeypots + machine learning , 2010, SIGIR.

[7]  Vanessa Murdock,et al.  Modeling locations with social media , 2013, Information Retrieval.

[8]  Tomoki Taniguchi,et al.  A Simple Scalable Neural Networks based Model for Geolocation Prediction in Twitter , 2016, NUT@COLING.

[9]  Aron Culotta,et al.  Inferring the origin locations of tweets with quantitative confidence , 2013, CSCW.

[10]  Heri Ramampiaro,et al.  Geo-temporal distribution of tag terms for event-related image retrieval , 2015, Inf. Process. Manag..

[11]  Yiming Yang,et al.  A Comparative Study on Feature Selection in Text Categorization , 1997, ICML.

[12]  B. Ripley Modelling Spatial Patterns , 1977 .

[13]  Wen Li,et al.  Geo-spatial Domain Expertise in Microblogs , 2014, ECIR.

[14]  Timothy Baldwin,et al.  Twitter Geolocation Prediction Shared Task of the 2016 Workshop on Noisy User-generated Text , 2016, NUT@COLING.

[15]  Mark Dredze,et al.  Geolocation for Twitter: Timing Matters , 2016, NAACL.

[16]  Themis Palpanas,et al.  Where has this tweet come from? Fast and fine-grained geolocalization of non-geotagged tweets , 2016, Social Network Analysis and Mining.

[17]  Aixin Sun,et al.  A Survey of Location Prediction on Twitter , 2017, IEEE Transactions on Knowledge and Data Engineering.

[18]  Timothy Baldwin,et al.  Text-Based Twitter User Geolocation Prediction , 2014, J. Artif. Intell. Res..

[19]  Geert-Jan Houben,et al.  Placing images on the world map: a microblog-based enrichment approach , 2012, SIGIR '12.