Some of you may recall that back in 2015 we surveyed our OCLC Research Library Partners to determine their top challenges with web archiving, and the need for guidance on metadata practices emerged as #1. In response, early in 2016 we established a Web Archiving Metadata Working Group (WAM) to develop best practices for metadata. The group did extensive background research over the past year, and we’re now on a fast track to publish three reports in the next several months. In the meantime, you can read a substantial overview of the project in this article published last Friday in the online Journal of Western Archives.
The first two reports will underpin the best practices: one on tools available for capture of websites, with a focus on their metadata extraction capabilities; and a review of the literature on metadata needs of web archives users.
The best practice guidelines will be in the third report. In addition to defining and interpreting a set of data elements, the report will articulate differences between bibliographic and archival standards; contrast approaches to description of individual websites and collections; and include both a literature review focused on metadata issues and crosswalks to related standards.
WAM established several principles to underpin the best practices. They are intended to …
- … address the needs of users of archived websites as determined by our literature review
- … be community-neutral, standards-neutral, and output-neutral; in other words, applicable to any context in which metadata for archived websites is needed
- … consist of a relatively lean set of data elements, with the scope of each defined (i.e., a data dictionary)
- … interpret each element for description of archived websites, which, unlike books or serials or published audiovisual media, have no conventions for representing elements such as creators, dates, or extent
- … be upward-compatible with standards that have far deeper data element sets, including RDA, MARC, DACS, EAD, and MODS
We are in the process of finalizing the set of data elements and have adopted the following so far:
These may seem both obvious and straightforward, but most need definition and interpretation for the web context. One example: what types of date are both feasible to determine and important to include, and how can their meaning be made clear? Additional elements under consideration include geographic coverage, publisher, rights, access, source of description, URL, and collector (or should the latter be owner? or repository? or location?). We’ve eliminated from consideration several that don’t have specific applicability to websites, including audience and statement of responsibility.
We’ll be circulating the draft best practices widely across the library and archives community and are hoping to hear from many who are struggling to describe websites and collections. Our aim is to promulgate best practices that will encourage use of metadata that is both meaningful and useful to users of these resources.