This research seeks to develop socio-computational approaches to improve citizen participation in rulemaking, using a pilot online participation platform, RegulationRoom. Rulemaking, the process through which federal regulatory agencies make new regulations, is an unusual method of federal decision-making because it formally incorporates a period of peer knowledge creation, the "notice-and-comment" process. Agencies proposing new rules must seek input from stakeholders, experts, and the general public, and consider all criticisms, questions, new data, and alternative ideas received. Yet only a limited segment of individuals and groups participate, and these participants engage in adversarial position-taking rather than collaborative deliberation. Neither of these shortcomings has been remedied by putting the process online. It is reasonable to hypothesize that broader, better public engagement requires purposefully designing e-rulemaking systems to provide norm enculturation and deliberation priming. This project will explore strategies for each, drawing on expertise in law, conflict resolution, natural language processing (NLP), machine learning, and recommender systems.

This research provides unique contributions in at least four areas: (a) advancing the state-of-the art in NLP by developing discourse analysis techniques to facilitate deliberative dialogue and collaborative knowledge production; (b) advancing recommender system and online community research to support mentoring activities and engagement with alternate points of view; (c) advancing the field of conflict resolution by adapting face-to-face moderation techniques to the online environment; and (d) extending legal understanding of how to increase transparency and participation by supporting broader, better public participation in rulemaking and other complex policymaking domains. It will also generate annotated datasets of comment and deliberation quality that will be released to other researchers. The results will extend the techniques and assessment measures available to moderators of online group discussion and provide new computational tools for improving the quality of individual contributions and for interpreting and synthesizing information during the knowledge production process.

Learning how to design effective Rulemaking 2.0 systems will strengthen the democratic process by expanding the range of individuals and groups who understand, engage with, and meaningfully contribute to federal policymaking. In addition, better policy outcomes should result from socio-computational techniques that enable meaningful participation by important but traditionally absent stakeholder groups, such as small business owners. State and local governments that use public comment processes will benefit similarly from this work as will non-governmental groups trying to increase effective participation in complex collaborative content creation online.

National Science Foundation (NSF)
Division of Information and Intelligent Systems (IIS)
Standard Grant (Standard)
Application #
Program Officer
William Bainbridge
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
Cornell University
United States
Zip Code