Over the past few years, a number of approximate inference algorithms for networked data have been put forth. We empirically compare the performance of three of the popular algorithms: loopy belief propagation, mean field relaxation labeling and iterative classification. We rate each algorithm in terms of its robustness to noise, both in attribute values and correlations across links. We also compare them across varying types of correlations across links.

@TechReport\{sen:um-tr07,
  author       = "Sen, Prithviraj and Getoor, Lise",
  title        = "Link-based Classification",
  institution  = "University of Maryland",
  number       = "CS-TR-4858",
  month        = "February",
  year         = "2007",
  type         = "Technical Report",
}