Exploiting geotagged resources for spatial clustering on social network services

Abstract

Nowadays, it has become common for users to geotag resources on many online social networking services. However, a large amount of data exists on social network services without annotations of their geographical location. Thus, it would be useful to tag these resources with geotags. This paper proposes a method to predict the location of unlabeled resources on social networking services. We use the Naive Bayes and support vector machine methods to classify the resources that are collected by using the term frequency of the tags in each class. In addition, we improve the calculation for these methods by using the values of the term frequency, and we invert the class frequency to optimize the input data. These results can be applied to tag unlabeled resources on social networking services.