In an effort to democratize the regulatory process, the United States Federal government created regulations.gov, a portal through which federal agencies can share proposed regulations and solicit feedback from the public. A proposed regulation will contain several requests for feedback on specific topics, and the public can then submit comments in response. While this reduces barriers to soliciting feedback, it still leaves regulators with a challenge: how to produce a summary and incorporate feedback from the sometimes tens of thousands of submitted comments. We propose an information retrieval system by which comments are aligned to specific regulatory requests. We evaluate several measures of semantic similarity for matching comments to information requests. We evaluate our proposed system over a dataset containing several regulations proposed for electronic cigarettes, an issue that energized tens of thousands of comments in response1.