Understanding Citizens' Direct Policy Suggestions to the Federal Government: A Natural Language Processing and Topic Modeling Approach

We report on our initial efforts to make sense of e-petitions as policy suggestions by using the NLP technique of \"topic modeling\" to identify the \"topics\" that emerge in e-petitions. Using a sample of petitions submitted to the Obama Administration's WtP petitioning system as a case study, we produced 30 emergent topics. 21 out of the 30 topics were initially coded as high-quality topics. Upon qualitative investigation, all but one of these 21 topics were determined to have a coherent theme. Our results imply that topic modeling has the potential to enable the interpretation of large quantities of citizen generated policy suggestions through a largely automated process, with potential application to research on e-participation and policy informatics.