Understanding Sensitivity: a First Step Towards Automating Sensitivity Review

Oliva, R. and Kim, Y. (2019) Understanding Sensitivity: a First Step Towards Automating Sensitivity Review. Archives, Access and AI: Working with Born-Digital and Digitised Archival Collections, London, UK, 15-17 Jan 2020.

Full text not currently available from Enlighten.

Abstract

Memory institutions face new challenges associated with curating data heritage in relation to “technological requirements for specialist skills, hardware, and software to render digital objects” (Harvey citation). In particular, it is generally accepted that manual approaches to processing digital records are becoming intractable, due to the volume of digital records stored by organisations (McDonald, Macdonald, and Ounis 2015). Sensitivity review is one of the necessary processes by which archivists determine which records may be released to the public, redacted, or closed to the public. This paper will discuss the complex nature of sensitivity review, for example in relation to legal mandates, controversial subjects, and cultural differences, to present the many factors that influence archivists in the sensitivity review process. Keeping a record open when it contains sensitive information can mean that memory institutions are breaking the law (Sloyan 2016), but a risk-averse approach such as restricting access to records that have not yet been reviewed may result in reduced levels of service for users and obscure subtle cultural dynamics involved in sensitivity. Lacking scalable approaches to tackle the complexity of sensitivity review and to mitigate such risks impedes archivists as they work to balance their responsibility to provide access to our data heritage with their duty to redact or close records that are sensitive. Some of these challenges could be addressed by incorporating automated or technology-assisted approaches. This paper will propose a nuanced understanding of sensitivity within archives and demonstrate that sensitive data is characterised by its context as much as its content, opening up a discussion regarding the potential of machine learning, information retrieval and natural language processing techniques in developing scalable technology-assisted sensitivity review workflows.

Item Type:Conference or Workshop Item
Status:Published
Refereed:Yes
Glasgow Author(s) Enlighten ID:Kim, Dr Yunhyong and Oliva, Rebecca
Authors: Oliva, R., and Kim, Y.
Subjects:Z Bibliography. Library Science. Information Resources > Z665 Library Science. Information Science
College/School:College of Arts & Humanities > School of Humanities > Information Studies
Related URLs:

University Staff: Request a correction | Enlighten Editors: Update this record