Using proxy servers in Iran, researchers Collin Anderson and Nima Nazeri scanned 800,000 Persian language Wikipedia articles. Every blocked article was identified and blocked pages were divided into ten categories to determine the type of content to which state censors are most adverse. In total, 963 blocked articles were found, covering a range of socio-political and sexual content including politics, journalism, the arts, religion, sex, sexuality, and human rights. Censors repeatedly targeted Wikipedia pages about government rivals, minority religious beliefs, and criticisms of the state, officials, and the police. The researchers found dozens of pages that seem to be unintentionally censored by keyword filtering, meaning that they were misidentified as sexual or profane and contained no content likely to offend Iranian authorities.