Smart Learning of Porn Fake News in the Family-Friendly Filters

¹ Department of Information and Communication, Tamkang University, Taipei, Taiwan
² National Taipei University of Technology, Taipei, Taiwan

^* Corresponding author: meilingj@mail.tku.edu.tw

Abstract

The proliferation of fake news on Facebook and Google has been a hot-button topic after the 2016 US presidential election. Fake news phenomenon is not limited in the political sphere. The porn industries have been using affiliate marketers to send fake news to reach more consumers, even children. Easy availability of pornography for children on the internet has been an issue. In US, the average age of exposure to porn is 11 to 12. Frequent exposure to pornography may lead to normalization of harmful behaviors. Starting late 2013, internet service providers in Britain made “family-friendly filters,” which block X-rated websites, the default for customers, because kids are exposed to pornography at a young age. Google banned pornographic ads from its search engine from July 2014. Prostitution and escort services extend its market despite these efforts for the sake of the upsurge porn fake news. Porn fake news is produced purposefully to click, share, react, and comment. To mitigate the damage caused by porn fake news, designing a fully automated fake news detector is currently infeasible, because the problem at hand is too complex for technology alone. Even the subproblem of defining the criteria under which to classify news as “fake” creates ambiguity that requires human judgment. The ability to determine whether an article is real or fake requires more than just information about the article; it requires an understanding of cultural factors, for example “tea” maybe used by prostitution and escort services in Taiwan. This paper suggests one way to use artificial intelligence and human judgment to make it more valid to quarantine porn fake news.