Sood, Sara, Judd Antin, and Elizabeth Churchill. "Profanity use in online communities." Proceedings of the SIGCHI conference on human factors in computing systems. 2012.

This paper discusses the challenges of detecting and managing inappropriate or objectionable content in user-generated web content, focusing on profanity detection systems. It highlights two major limitations of current list-based profanity detection approaches: their inability to adapt to the evolving nature of language, including misspellings, abbreviations, and new slang, and their lack of customization for different communities and contexts, which have unique norms and tolerances for what is considered acceptable. By analyzing comments from a social news site, the study demonstrates the inadequacy of existing systems and suggests that a more effective approach to profanity detection would consider the specific context and community norms, moving towards systems that are capable of understanding and adapting to the nuances of language and social settings.