Debunking Intermediary Censorship Framework in Social Media via a Content Retrieval and Classification Software

Following the enactment of computer crime law in Thailand, online service providers are compelled to control illegal content including content that is deemed harmful or problematic. This situation leads to self-censorship of intermediaries, often resulting in overblocking to avoid violating the law. Such filtering flaw both infringes users’ freedom of expression and impedes the business of OSPs in Thailand. The Innovative Retrieval System (IRS) is thus developed to investigate intermediary censorship in online discussion forum, Pantip.com, as a case study of social media. The result shows that there is no consistency of censorship pattern on the website at all. The censorship criteria depend on type of content in each forum. Overblocking is also high, over 70% of removed content, due to intimidation of governmental agencies, lawsuits from business organizations, and fear of intermediary liability. Website administrator admitted that he would cut off some users to avoid business troubles.