Data Curation Issues in the Chemical Sciences

Abstract

Given the current emphasis on the diligent management of research data, it is almost self-evident that curation should be a fundamental aspect of the research practice in the chemical sciences. An essential ingredient in the curation process is metadata: descriptive information and classification labels that group related items, provide context and facilitate the reuse of specified research outputs. In our view, the single most important function of metadata is to capture context. Problems can and do arise later in the research cycle if researchers do not capture the correct context as they record their experiments and acquire their data, so curation is naturally most effective when researchers capture metadata at source. However, researchers perceive curation as a burden, raising a range of issues that we examine generically and with regard specifically to the challenges that the burden of creation presents in the chemical sciences. We examine the significance of training in data management and explore the potential role for librarians. We also consider the issues relating to the deployment of controlled vocabularies. We conclude with the view that chemists have yet to take full advantage of the extra opportunities created by effective curation, emphasising the importance of capturing contextual information at the earliest possible stage: at source.