AI-powered narrative building for facilitating public participation and engagement

Fernando Marmolejo-Ramos, Thomas Workman, Clint Walker, Don Lenihan, Sarah Moulds, Juan C Correa, Anca M Hanea, Belona Sonna

Research output: Contribution to journalArticlepeer-review


Algorithms, data, and AI (ADA) technologies permeate most societies worldwide because of their proven benefits in different areas of life. Governments are the entities in charge of harnessing the benefits of ADA technologies above and beyond providing government services digitally. ADA technologies have the potential to transform the way governments develop and deliver services to citizens, and the way citizens engage with their governments. Conventional public engagement strategies employed by governments have limited both the quality and diversity of deliberation between the citizen and their governments, and the potential for ADA technologies to be employed to improve the experience for both governments and the citizens they serve. In this article we argue that ADA technologies can improve the quality, scope, and reach of public engagement by governments, particularly when coupled with other strategies to ensure legitimacy and accessibility among a broad range of communities and other stakeholders. In particular, we explore the role “narrative building” (NB) can play in facilitating public engagement through the use of ADA technologies. We describe a theoretical implementation of NB enhanced by adding natural language processing, expert knowledge elicitation, and semantic differential rating scales capabilities to increase gains in scale and reach. The theoretical implementation focuses on the public’s opinion on ADA-related technologies, and it derives implications for ethical governance.
Original languageEnglish
JournalDiscover Artificial Intelligence
Issue number7
Publication statusPublished - 2022
Externally publishedYes


Dive into the research topics of 'AI-powered narrative building for facilitating public participation and engagement'. Together they form a unique fingerprint.

Cite this