Systems for Collective Human Curation of Online Discussion

The internet was supposed to democratize discussion, allowing people from all walks of life to communicate with each other at scale. However, this vision has not been fully realized--instead online discourse seems to be getting worse, as people are increasingly drowning in discussion, with much of it unwanted or unpleasant. In this thesis, I present new systems that empower discussion participants to work collectively to bring order to discussions through a range of curation tools that superimpose richer metadata structure on top of standard discussion formats. These systems enable the following new capabilities: 1) recursive summarization of threaded forums using Wikum, 2) teamsourced tagging and summarization of group chat using Tilda, 3) fine-grained customization of email delivery within mailing lists using Murmur, and 4) friendsourced moderation of messages against online harassment using Squadbox.

In a world of abundant discussion and mass capabilities for amplification, the curation of a social space becomes as equally essential as content creation in defining the nature of that space. By putting more powerful techniques for curation in the hands of everyday people, I envision a future where end users are empowered to actively co-create every aspect of their online discussion environments, bringing in their nuanced and contextual insights.