A computational model for corruption assessment

Abstract

Corruption afflicts public services world-wide and is universally considered undesirable. However, there is little prior work to formalize it so that it can be attacked effectively using computational (ICT) techniques. It is believed that trends like process (service) automation and open data help tackle corruption but little is known about why and when. In this work, we address this gap by (a) presenting a meta-model for agent-delivered services motivated by real case studies in corruption, (b) describing common patterns of corruption in such service process designs, (c) formalizing the patterns via a language that instantiates the meta-model, and (d) evaluating the formal patterns on new cases of corruption. We then discuss how the formal patterns help evaluate effectiveness of the various mechanisms for addressing corruption. © 2013 ACM.