Human-Data Interaction: Thinking beyond individual datasets

TL;DR

This paper emphasizes the importance of viewing data as an interactive process rather than static resources, advocating for Human Data Interactions to enhance data usability and social engagement.

Contribution

It introduces the concept of Human Data Interactions, highlighting the need to consider data as a dynamic process with interactional affordances, supported by a case study of Kaggle.

Findings

Data accessibility alone is insufficient for usability.

Viewing data as an interactive process opens new possibilities.

Human Data Interactions can improve engagement and understanding.

Abstract

Having greater access to data leads to many benefits, from advancing science to promoting accountability in government to boosting innovation. However, merely providing data access does not make data easy to use; even when data is openly available online, people may struggle to work with it. In this article, we draw on prior work, including our own, and a case study of Kaggle (a large online data science community) to discuss the importance of moving away from viewing datasets as static resources. Instead, we describe the view of data as a process with its own interactional affordances that offer many different possibilities for data, as well as for social interaction. We advocate for the notion of Human Data Interactions and their potential implications for various audiences.