Research data can be viewed as scholarly products to be shared, mined, and curated – or as transient artifacts of the research process, incomprehensible without context, domain knowledge, and associated software and technologies. These two perspectives represent but one of many dimensions over which research data vary. Data are not things. Rather, they are representations of observations, objects, or other entities used as evidence of phenomena for the purposes of research or scholarship. They are created in a context, often as part of collaborative research activities. Whereas data practices vary widely across domains, disciplines, and cultures, research policy and technological infrastructures promote uniform approaches to data management. By celebrating the diversity of research data, their value and richness may be enhanced. However, that diversity poses challenges for preserving context, for stewardship, for exploiting data in collaborations across research domains, and for reuse over the short and long term.
Available at: http://works.bepress.com/borgman/354/