Tuesday, July 07, 2015

Collection, Curation, Citation at Source: Publication@Source 10 Years On

Collection, Curation, Citation at Source: Publication@Source 10 Years On. Jeremy G. Frey, et al. International Journal of Digital Curation. Vol 10, No 2, 2015.
   The article describes a scholarly knowledge cycle which says the accumulation of knowledge is based on the continuing use and reuse of data and information. Collection, curation, and citation are three processes intrinsic to the workflows of the cycle. The currency of collection, curation, and citation is metadata."Policies should recognize that small amounts of adequately characterized, focused data are preferable to large amounts of inadequately defined and controlled data stored in a random repository." The increasing size of data-sets and the growing risk of loss through catastrophic failure (such as a disk failure) has led to researchers to use cloud storage, perhaps too uncritically so.

The responsibilities of researchers for meeting the requirements of sound governance and ensuring the quality of their work have become more apparent. The article places the responsibility for curation firmly with the originator of the data. "Researchers should organize their data and preserve it with semantically rich metadata, captured at source, to provide short- and long-term advantages for sharing and collaboration."  Principal Investigators, as custodians, are particularly responsible for clinical data management and security (though curation and preservation activities exist in other research roles). "Curators usually attempt to add links to the original publications or source databases, but in practice, provenance records are often absent, incomplete or ad hoc, often despite curators’ best efforts. Also, manually managed provenance records are at higher risk of human error or falsification." There is a pressing need for training and education to encourage researchers to curate the data as they collect it at source.

"All science is strongly dependent on preserving, maintaining, and adding value to the research record, including the data, both raw and derived, generated during the scientific process. This statement leads naturally to the assertion that all science is strongly dependent on curation."

No comments: