Placing User-generated Content on the Map with Confidence

Placing User-generated Content on the Map with Confidence. Intagorn, S. & Lerman, K. In ACM GIS, 2014.

We describe a method that predicts the location of user-generated content using textual features alone. Unlike previous methods for geotagging text documents, our proposed method is not sensitive to how we discretize space. We also discover that spatial resolu- tion has an impact on the prediction accuracy, which allows us to trade-off the spatial resolution of the predicted location against our confidence about its accuracy. Our method can be used to estimate the error in document�s predicted location, enabling us to filter out poor quality predictions. We evaluate the proposed method exten- sively on user-generated content collected from two different social media sites, Flickr and Twitter. Our evaluation examines its perfor- mance on the geotagging task and with respect to different parame- ters. We achieve state-of-the-art results for all three tasks: location prediction, error estimation and result ranking and also provide a theoretical explanation of the effect of spatial resolution factor on geotagging accuracy. Our findings provide valuable insights into the design of geotagging systems and their quality control.

@INPROCEEDINGS{Intagorn14acmgis,
  AUTHOR =       {Suradej Intagorn and Kristina Lerman},
  TITLE =        {Placing User-generated Content on the Map with Confidence},
  BOOKTITLE =    {ACM GIS},
  YEAR =         {2014},
  abstract =     {We describe a method that predicts the location of user-generated
content using textual features alone. Unlike previous methods for
geotagging text documents, our proposed method is not sensitive
to how we discretize space. We also discover that spatial resolu-
tion has an impact on the prediction accuracy, which allows us to
trade-off the spatial resolution of the predicted location against our
confidence about its accuracy. Our method can be used to estimate
the error in document�s predicted location, enabling us to filter out
poor quality predictions. We evaluate the proposed method exten-
sively on user-generated content collected from two different social
media sites, Flickr and Twitter. Our evaluation examines its perfor-
mance on the geotagging task and with respect to different parame-
ters. We achieve state-of-the-art results for all three tasks: location
prediction, error estimation and result ranking and also provide a
theoretical explanation of the effect of spatial resolution factor on
geotagging accuracy. Our findings provide valuable insights into
the design of geotagging systems and their quality control.},
  keywords =     {social-annotation},
  url={http://www.isi.edu/integration/people/lerman/papers/Intagorn14acmgis.pdf}
}

Downloads: 2