EAGER: Investigating Diversity in Online Community Filtering
Drexel University, Philadelphia PA
Investigators
Abstract
A transition is occurring from a world in which gatekeepers and editors filter content before it is published to a world full of user-generated content in which information filtering is done after publication. Today's online communities have developed a variety of community-based filtering and rating mechanisms to help maintain quality and manageability. However, it is an open question whether these filtering mechanisms represent "the wisdom of crowds" or "the censoring mob." This project will apply statistical machine learning and ethnographic studies to understand the mechanisms by which online communities censor content from the bottom up. This understanding will provide insight into how values are and can be embedded into these large, socially intelligent systems. Ultimately, the goal is to design socially intelligent community filtering systems in which individuals, communities, and intelligent software agents collaborate, to explain the mechanisms behind social, bottom-up filtering, and expand the range of the possible in terms of the values these systems can reflect and the communities it can serve. This project will study the mechanisms through which the social construction of gender impacts community filtering systems. This will be done via an in-depth study of two online communities that have vigorous community policed comment filtering; one whose participants are predominantly male and another whose participants are predominantly female. Online communities are rapidly becoming the modern public square and community filtering has the potential to make the space vibrant and useful and/or degenerate into a form of censorship. The health of our civil society and its ability to address large challenges depends on the health of its public discourse. By creating systems for socially intelligent filtering that reflect the community we facilitate diversity, in that minority positions are protected and preserved, while at the same time majority positions have the opportunity to develop and refine cogent arguments necessary for a well reasoned debate.
View original record on NSF Award Search →