Determining the Provenance of User Decisions

18^th March 2014, 16:00 Ashton Lecture Theater
Dr Simon Miles
Agents and Intelligent Systems
Department of Informatics
King's College London

Abstract

People are involved somewhere within most software processes, making decisions that affect how the processes then continue. In a large organisation, processes can involve many people with differing aims depending on their roles. This means that to understand the history of the outputs of such processes, i.e. the provenance of the outputs, we need to establish the basis on which decisions were made. In general, understanding the provenance of data is important in many domains, for knowing how to interpret the data and for determining whether to rely on it. In this talk, I will discuss technologies that may be used in tracking provenance involving human decisions: how to model provenance in W3C PROV, how to engineer applications to capture relevant provenance information, and how we might compare simulations of human decision making with provenance records to determine plausible explanations of decisions. The ideas will be illustrated in the context of a clinical research project.​