Dissertation Title:

Towards Bridging and Governing Decentralized Communities: Insights and Tools for Constructive Discourse

Abstract: 

"Unless the spaces in a building are arranged in a sequence which corresponds to their degrees of privacy the visits made by strangers, friends, guests, clients, family, will always be a little awkward." (Alexander, 1977) ––– Unlike physical spaces where we have the capacity to move fluently between various environments, much of our online experience takes place in loud, crowded, and imposed public areas. This often hinders meaningful connections and understanding among individuals and deepens divides. Further, as our social spheres or communities–with differing norms and values–coincide in the same spaces friction and misunderstanding lead to toxic behavior that is more often than not monitored and governed at the platform level rather than at the community level, approaching a values problem with a one-size-fits-all type solution.

Fortunately, new strategies are emerging. The surge of decentralized protocols and networks has started to shift this schema to a more nuanced governance schema. Yet, these approaches remain highly technical and out of reach for most social media users. Creating healthier, intentional spaces, which involve mediating boundaries, social norms, and values with focus on increasing user agency on their community governance and designing for non-tech-expert users is a crucial step towards addressing these challenge.

This thesis aims to establish that designing with a focus in bottom-up and decentralized governance rises opportunities that come as an after thought when social networks are designed without purpose. Further, this theses looks at the intentional design of online spaces called communities, along with bridges between the, which we  argue can yield benefits in fostering healthier conversations. On the other hand, because our online experience is largely shaped by content that is filtered and ranked by algorithms according to centralized guidelines, we intend to develop the knowledge necessary to advance content moderation strategies and governability for constructive communication across decentralized communities.

This dissertation collects and analyzes historical social media data to identify the drivers of pro-social behaviors within online communities. Subsequently, it designs and test tools that allow communities to govern their interactions through content moderation aligned with their own speech norms. Additionally, it proposes new ways to enable distinct communities to bridge their social expectations through understanding their differences, assisted by large language models.

In part one, we introduce and explore a new dataset–the largest to our knowledge–focused on community-centered content moderation from Reddit, where each community has set its moderation guidelines. We recovered over 230K removed posts, which are no longer available for public access. We examine over 19K communities defined by mission and over 60K speech norms and propose an empirically grounded norms schema for self-governed communities. We observe the relationship between pro-social behavior and community purpose, revealing a correlation between a community's mission statement and the quality of conversation. Designing for communities to self-govern with their speech norms presents the challenge of shifting from a centralized, top-down approach to a distributed approach. While centralized methods offer efficiency in affordances when using highly accurate language models, they limit community members' participation in shaping how content is curated and ranked within their community. In part two, we prototype tools for i) explainable decentralized content moderation, and for ii) bridging communities with differing speech norm. Finally, in part three, we consolidate these learnings and introduce a new social network space, Odessa. Odessa is a decentralized social network foundationally designed for user-friendly decentralized governance, with a focus on bridging opportunities between distinct communities. Our design process, user experiments, and evaluation schema reveal insights and opportunities for the future of decentralized social media governance.

Deb K. Roy
Professor of Media Arts and Sciences
Massachusetts Institute of Technology

Rosalind W. Picard
Professor of Media Arts and Sciences
Massachusetts Institute of Technology

Jonathan L. Zittrain
George Bemis Professor of International Law
Harvard University