Discovering and Characterizing Places of Interest Using Flickr and Twitter

Discovering and Characterizing Places of Interest Using Flickr and Twitter

Steven Van Canneyt (iMinds, Department of Information Technology (INTEC), Ghent University, Gent, Belgium), Steven Schockaert (School of Computer Science & Informatics, Cardiff University, Cardiff, United Kingdom) and Bart Dhoedt (iMinds, Department of Information Technology (INTEC), Ghent University, Gent, Belgium)
Copyright: © 2013 |Pages: 28
DOI: 10.4018/ijswis.2013070105
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

Databases of places have become increasingly popular to identify places of a given type that are close to a user-specified location. As it is important for these systems to use an up-to-date database with a broad coverage, there is a need for techniques that are capable of expanding place databases in an automated way. In this paper the authors discuss how geographically annotated information obtained from social media can be used to discover new places. In particular, the authors first determine potential places of interest by clustering the locations where Flickr photos have been taken. The tags from the Flickr photos and the terms of the Twitter messages posted in the vicinity of the obtained candidate places of interest are then used to rank them based on the likelihood that they belong to a given type. For several place types, their methodology finds places that are not yet contained in the databases used by Foursquare, Google, LinkedGeoData and Geonames. Furthermore, the authors’ experimental results show that the proposed method can successfully identify errors in existing place databases such as Foursquare.
Article Preview

Introduction

Berners-Lee’s vision of the Semantic Web (Berners-Lee, 2001) has become increasingly popular in the last few years. The World Wide Web would evolve to a highly interconnected network of data that could be easily accessed and understood by machines. Applications could for instance use the Semantic Web to construct customized answers to a particular question. In such applications the user is no longer required to search for information or pore through results. A question can e.g. be ‘What are locations of the restaurants in London?’. To answer this question, a structured dataset has to be available containing places located in London (entities), associated with their location and semantic type (properties).

However, a lot of information on the Web is still unstructured or only semi-structured. Therefore, there is a need for automated methods to extend structured datasets using existing Web data. Several methods of this form have been proposed, e.g. YAGO2 (Hoffart, 2013) and BabelNet (Navigli, 2010) are knowledge bases that are constructed using Wikipedia and Wordnet. Other research focuses on establishing structured datasets containing information of a specific type. For instance, LinkedGeoData (Auer, 2009) is a dataset of places constructed using OpenStreetMap, an application in which users can submit geographical data such as place semantics.

In this paper, we will focus on improving existing databases of places. More precisely, we will add new places and discover likely errors using data from the Web. Social media data is particularly promising in this respect, due to the large amounts of geographically annotated data produced by these media. For example, about 1.5% of all Twitter posts (i.e. tweets) are annotated with geographical coordinates (Murdock, 2011). In addition, there are currently more than 190 million geotagged Flickr photos (Flickr, 2013). This data has been used to e.g. automatically detect events (Rattenbury, 2007; Sakaki, 2010; Lee, 2011), to find popular places (Crandall, 2009; Van Canneyt, 2011) and tourist routes (Choudhury, 2010; Jain, 2010).

Complete Article List

Search this Journal:
Reset
Open Access Articles
Volume 13: 4 Issues (2017)
Volume 12: 4 Issues (2016)
Volume 11: 4 Issues (2015)
Volume 10: 4 Issues (2014)
Volume 9: 4 Issues (2013)
Volume 8: 4 Issues (2012)
Volume 7: 4 Issues (2011)
Volume 6: 4 Issues (2010)
Volume 5: 4 Issues (2009)
Volume 4: 4 Issues (2008)
Volume 3: 4 Issues (2007)
Volume 2: 4 Issues (2006)
Volume 1: 4 Issues (2005)
View Complete Journal Contents Listing