Human-Data Interaction in Healthcare

Background

As rightly noted by Hornung et al. (2015), the expression HDI is neither specific nor new. On the one hand, humans have always interacted with data, that is they have inscribed, read and communicated data in terms of simple marks, more complex symbols and any kind of meaningful signs¹. On the other hand, this interaction has predominantly been mono-directional until interactive systems have allowed users to get access to data according to their peculiar and situated needs. In the computing literature, the expression has been used since the 1990s, e.g., by Kennedy et al. (1996), to address the needs of the users in the context of data intensive settings and hence “the problem of delivering personalized, context-aware, and understandable data from big datasets” (Cafaro, 2012). More recently a number of authors have tried to define Human-Data Interaction, by focusing on different aspects: Elmqvist (2011) and Cafaro (2012) focused on the material and embodied aspects of interaction; Crabtree and Mortier (2015) prefer to assimilate HDI to “the management and use of personal data in society at large” where personal here means that data are either “about individuals” or are delivered in terms of “personalized experiences”. In particular, Elmqvist (2011) defines HDI as “the human manipulation, analysis, and sense making of large, unstructured, and complex datasets” and Cafaro sees HDI systems “as technologies that use embodied interaction to facilitate the users' exploration of rich datasets”. We agree with Hornung et al. (2015) and Crabtree and Mortier (2014) in regard to the fact that “HDI does not only refer to embodied interaction, but to all kinds of interaction” and that “HDI should investigate ‘data that affects people’ ”.