Font Size:
Making Metadata Perform: Metadata Creation from Past, Present and Future Projects
Heather Leasor, Janet McDougall, Steven McEachern
Building: Holme Building
Room: Sutherland Room
Date: 2016-07-22 11:00 AM – 12:30 PM
Last modified: 2016-05-06
Abstract
The Australian Data Archive has a diverse range of clientele who deposit data and metadata with varying degrees of breadth, depth, complexity and completeness. For the archive, we regularly need to consider how we can support our depositors to generate high quality data and metadata.
Methods for metadata creation, for both archival processes and researcher practice, are the focus of this paper. Building metadata collection into the research process potentially allows reuse of metadata within the research organisation outside standard archive use, such as building reports for clients, internal monitoring, and documentation. There are also corollary benefits for the archive in the efficient and accurate generation of project metadata.
To understand these benefits and how they might be realised, we first need to understand existing practices. We begin with a comparison of two existing collections:
- a historical crime data set from a government agency collected 15 years prior to deposit of the data, which came with limited documentation, but with access to public domain information, and to some of the members of the original research team.
- a recently collected dataset deposited by an Australian data collection agency, accompanied by broad documentation, technical information and access to the researchers. We detail in these cases the issues around developing metadata at the archiving stage from past resources, challenges in determining the "source of truth", and associated practical concerns.
We then turn to the question of how we can learn from the past in supporting metadata creation as part of the planning process for future studies? We argue, on the basis of our cases, that the most effective mechanism for metadata creation is at the time of the research activities take place. As such, we highlight the importance of building metadata collection into the research process (in this case using DDI 2.5).
The final section of the paper then explores the potential for embedding metadata collection at the time of data creation. Here we again use a current case study where ADA is working with the data collection agency, to establish practices for capturing metadata (either explicitly or through unobtrusive systems). In particular, this case study considers how DDI and associated metadata standards can be used to help the agency to best utilize metadata for their own internal purposes, and subsequently to assist the agency and ADA in enabling more timely and efficient data archiving and effective secondary data use.
Methods for metadata creation, for both archival processes and researcher practice, are the focus of this paper. Building metadata collection into the research process potentially allows reuse of metadata within the research organisation outside standard archive use, such as building reports for clients, internal monitoring, and documentation. There are also corollary benefits for the archive in the efficient and accurate generation of project metadata.
To understand these benefits and how they might be realised, we first need to understand existing practices. We begin with a comparison of two existing collections:
- a historical crime data set from a government agency collected 15 years prior to deposit of the data, which came with limited documentation, but with access to public domain information, and to some of the members of the original research team.
- a recently collected dataset deposited by an Australian data collection agency, accompanied by broad documentation, technical information and access to the researchers. We detail in these cases the issues around developing metadata at the archiving stage from past resources, challenges in determining the "source of truth", and associated practical concerns.
We then turn to the question of how we can learn from the past in supporting metadata creation as part of the planning process for future studies? We argue, on the basis of our cases, that the most effective mechanism for metadata creation is at the time of the research activities take place. As such, we highlight the importance of building metadata collection into the research process (in this case using DDI 2.5).
The final section of the paper then explores the potential for embedding metadata collection at the time of data creation. Here we again use a current case study where ADA is working with the data collection agency, to establish practices for capturing metadata (either explicitly or through unobtrusive systems). In particular, this case study considers how DDI and associated metadata standards can be used to help the agency to best utilize metadata for their own internal purposes, and subsequently to assist the agency and ADA in enabling more timely and efficient data archiving and effective secondary data use.