The Need for Sensemaking in Networked Privacy and Algorithmic Responsibility
Max Van Kleek‚ William Seymour‚ Michael Veale‚ Reuben Binns and Nigel Shadbolt
This paper proposes that two significant and emerging problems facing our connected, data-driven society may be more effectively solved by being framed as sensemaking challenges. The first is in empowering individuals to take control of their privacy, in device-rich information environments where personal information is fed transparently to complex networks of information brokers. Although sensemaking is often framed as an analytical activity undertaken by experts, due to the fact that non-specialist end-users are now being forced to make expert-like decisions in complex information environments, we argue that it is both appropriate and important to consider sensemaking challenges in this context. The second is in supporting human-in-the-loop algorithmic decision-making, in which important decisions bringing direct consequences for individuals, or indirect consequences for groups, are made with the support of data-driven algorithmic systems. In both privacy and algorithmic decision-making, framing the problems as sensemaking challenges acknowledges complex and ill- defined problem structures, and affords the opportunity to view these activities as both building up relevant expertise schemas over time, and being driven potentially by recognition-primed decision making.