Many scholars believe that electronic rulemaking has great but largely untapped potential to expand the public's democratic input and improve federal agency regulatory rules. The existing federal rulemaking process, however, elicits many redundant and poorly considered comments, and some participants challenge the legitimacy of agency rulemaking decisions. This Deliberative ERulemaking (DeER) project tests the application of Natural Language Processing (NLP) technologies and public deliberation techniques to improve the quality and organization of public comments and the legitimacy of the federal rulemaking process. This paper describes the goals, innovations, research design, technology, and preliminary results of the DeER project. The social science herein advances research on measuring the quality of public deliberation, the effects of deliberation on individuals and communities, and the emergence of distributed intelligence in networked decision making groups. Categories and ...