{"id":1050,"date":"2010-03-31T10:44:47","date_gmt":"2010-03-31T09:44:47","guid":{"rendered":"http:\/\/blog.soton.ac.uk\/keepit\/?p=1050"},"modified":"2010-03-31T10:44:47","modified_gmt":"2010-03-31T09:44:47","slug":"digital-preservation-tools-for-repository-managers-primer-on-preservation-workflow-formats-and-characterisation","status":"publish","type":"post","link":"https:\/\/generic.wordpress.soton.ac.uk\/test-media\/2010\/03\/31\/digital-preservation-tools-for-repository-managers-primer-on-preservation-workflow-formats-and-characterisation\/","title":{"rendered":"KeepIt course 3: Primer on preservation workflow, formats and characterisation"},"content":{"rendered":"

KeepIt course module 3, London, 2 March 2010<\/em>
\n Tools<\/strong> this module: Significant properties<\/a>, PREMIS<\/a>, Open Provenance Model<\/a> (OPM)
\nTags<\/strong> Find out more about: this module KeepIt course 3<\/a>, the full KeepIt course<\/a>
\nPresentations and tutorial exercises<\/strong> course 3<\/a> (source files)<\/p>\n

\"Typical<\/a>This post was updated on 9 April 2010.<\/strong><\/p>\n

In this module we really engaged with issues right at the core of preservation \u2013 developing an understanding of the real properties, functions, behaviours, structures, content and contexts which together provide a thorough understanding of a digital object. It is critical to have this understanding and clarity before we can be confident that we are working to preserve what\u00a0truly is important about an \u201cobject\u201d.<\/p>\n

The module focussed on:<\/p>\n

    \n
  1. Preservation workflow and format risks<\/li>\n
  2. Significant properties<\/li>\n
  3. Preservation metadata and provenance<\/li>\n<\/ol>\n

    This was designed to act as a primer for the following KeepIt course 4, which put this into practice using a preservation planning tool and repository applications.<\/p>\n

    1 Preservation workflow and format risks<\/h2>\n

    Steve Hitchcock, KeepIt project,\u00a0University of Southampton<\/p>\n

    We started our day with a short, informal and enjoyable game for sharing significant data and characteristics amongst small groups. Different groups were given simple numeric data, or four playing cards, and asked to transmit the data around the group using a given frequency of transcription. Sometimes the data was preserved (numbers or value of playing cards), sometimes the format (suit and\/or colour) and sequence, sometimes both; and, as we discovered, sometimes data is lost or errors are introduced, and sometimes these can be corrected. This gave us a useful way to capture our attention for significant properties and the actions involved in preserving data.<\/p>\n

    First, some important background terms necessary to understand the concept of significant properties were introduced.<\/p>\n

    Open Archival Information System (OAIS) reference model:
    \nData object interpreted via <\/strong>Representation Information yields<\/strong> Information Object
    \nNational Archives of Australia (NAA) Performance model:
    \nSource interpreted via <\/strong>Process yields<\/strong> Performance<\/p>\n

    OAIS, as a reference model, provides a way to compare digital preservation systems. This provides us with a mechanism to establish trust between different approaches. It also\u00a0provides a model that can be aligned with repository processes, such as deposit, content management and access. From an OAIS and content management perspective this can be divided further into preservation-related processes, and among these we find three to describe our preservation workflow for file format management (Table 1).<\/p>\n\n\n\n\n
    \n

    Check<\/strong><\/p>\n<\/td>\n

    		\n

    Analyse<\/strong><\/p>\n<\/td>\n

    		\n

    Action<\/strong><\/p>\n<\/td>\n<\/tr>\n

    Format \u2013 version, verification (tools available: JHOVE, DROID)<\/td>\nPreservation planning \u2013Significant properties, provenance, technical characteristics, risk factors, (PLATO, PRONOM, Inform)<\/td>\nMigration, emulation, storage selection<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

    Table 1. Preservation workflow for file format management<\/strong><\/p>\n

    We were given a practical exercise to make a comparison between one format over another, using the PRONOM inherent properties of file formats, and decide which performs better. Groups were given a free choice of which two formats to compare, and each group chose a different comparison. No prior knowledge of the properties was assumed, other than the simple descriptions provided, or they were assumed to be self-descriptive. Two properties were excluded as these are less easy to evaluate without some prior knowledge.<\/p>\n\n\n\n\n\n\n\n\n\n\n\n\n\n
    \n

    PRONOM Inherent Property<\/p>\n<\/td>\n

    		\n

    Word\/PDF<\/p>\n<\/td>\n

    		\n

    TIFF\/JPEG<\/p>\n<\/td>\n

    		\n

    PDF\/XML<\/p>\n<\/td>\n<\/tr>\n

    1000 Ubiquity<\/td>\n\n

    PDF<\/p>\n<\/td>\n

    		\n

    JPEG<\/p>\n<\/td>\n

    		\n

    PDF<\/p>\n<\/td>\n<\/tr>\n

    1001 Support<\/td>\n\n

    PDF<\/p>\n<\/td>\n

    		\n

    =<\/p>\n<\/td>\n

    		\n

    XML<\/p>\n<\/td>\n<\/tr>\n

    1002 Disclosure<\/span><\/td>\n<\/td>\n<\/td>\n<\/td>\n<\/tr>\n
    1003 Document<\/span><\/td>\n<\/td>\n<\/td>\n<\/td>\n<\/tr>\n
    1004 Stability<\/td>\n\n

    =<\/p>\n<\/td>\n

    		\n

    TIFF<\/p>\n<\/td>\n

    		\n

    XML<\/p>\n<\/td>\n<\/tr>\n

    1005 Ease of identification<\/td>\n\n

    = (or marg.PDF)<\/p>\n<\/td>\n

    		\n

    =<\/p>\n<\/td>\n

    		\n

    =<\/p>\n<\/td>\n<\/tr>\n

    1006 Ease of validation<\/td>\n\n

    PDF (internal mechanisms)<\/p>\n<\/td>\n

    		\n

    =<\/p>\n<\/td>\n

    		\n

    XML<\/p>\n<\/td>\n<\/tr>\n

    1007 Lossiness<\/td>\n\n

    =<\/p>\n<\/td>\n

    		\n

    TIFF<\/p>\n<\/td>\n

    		\n

    XML<\/p>\n<\/td>\n<\/tr>\n

    1008 IP<\/td>\n\n

    =<\/p>\n<\/td>\n

    		\n

    =<\/p>\n<\/td>\n

    		\n

    XML<\/p>\n<\/td>\n<\/tr>\n

    1009 Complexity<\/td>\n\n

    Word<\/p>\n<\/td>\n

    		\n

    TIFF<\/p>\n<\/td>\n

    		\n

    XML<\/p>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

    Table 2. \u00a0Comparing popular formats with reference to file format properties<\/strong><\/p>\n

    As can be seen from Table 2, in each case a clear format winner was identified based on the analysis provided.<\/p>\n

    We were then asked to consider why we might choose NOT to use the format that performed better for these criteria<\/em><\/strong>:<\/strong>
    \n\u2022\tPDF\/Word – Why not PDF? PDF is essentially a conversion format, not a source authoring format.
    \n\u2022\tTIFF\/JPEG – Why not TIFF? JPEG is compressed, would take\u00a0up less space in storage. This factor may be crucial. Archival\u00a0quality copy or a derivative?
    \n\u2022\tXML\/PDF – Why not XML? Many repository resources are deposited in\u00a0PDF. Do people understand what they need to do with XML?<\/p>\n

    Some thoughts about formats [1, 2]:
    \n\u2022\tFree vs open source vs open standard?
    \n\u2022\tMS Office \u2013 XML \u2013 open standard (Word doc can be saved as\u00a0XML)
    \n\u2022\tOpen Office \u2013 free \u2013 XML \u2013 open standard
    \n\u2022\tPDF \u2013\u00a0page representation
    \n\u2022\tXML \u2013\u00a0generic web format, computational<\/p>\n

    We recognised that it is crucial to work with content creators for repositories \u2013 authors, typically, cannot be expected to check to see if they are required to provide a converted copy of a resource [3].<\/p>\n

    Steve Hitchcock summarised by observing that the issue is essentially that of risk assessment \u2013 if we had identified a risk in our personal lives, we would wish to have some way to moderate or to manage the risk \u2013 by means of an insurance policy, for instance…..or smoke detectors, alarm systems etc. For repository content there may be very specific risks which we need to undertake detailed analysis and provide specific solutions.<\/p>\n

    References:
    \n1 Repositories Support Project briefing document on Preservation & Storage Formats for Repositories(May 2008)\u00a0http:\/\/www.rsp.ac.uk\/pubs\/briefingpapers-docs\/technical-preservformats.pdf<\/a><\/p>\n

    2 Rosenthal, D., dshr\u2019s blog, accessed 24 March 2010, various posts on file formats, e.g.\u00a0Are format specifications important for preservation? (January 4, 2009),\u00a0Format Obsolescence: the Prostate Cancer of Preservation (May 7, 2007),\u00a0Format Obsolescence: Scenarios (April 29, 2007) http:\/\/blog.dshr.org\/<\/a><\/p>\n

    3 Ashby, S., Summary of responses to IR questionnaire. JISC-Repositories, 18 February 2010\u00a0[online]. Available from: JISC-REPOSITORIES@JISCMAIL.AC.UK [Accessed 24 March 2010]\u00a0http:\/\/bit.ly\/8Zqdjl<\/a><\/p>\n

    2 Introduction to Significant Properties<\/h2>\n

    Stephen Grace and Gareth Knight from Kings College London.
    \n    http:\/\/www.significantproperties.org.uk<\/a><\/p>\n

    Stephen Grace started off this detailed section of the day with an introduction to the understanding and definition of Significant Properties and their relevance to work of perservation of digital resources. The InSPECT Project: \u201c…adapted the Function-Behaviour-Structure (FBS) framework, a framework developed by John Gero to assist engineers and designers with the process of creating and re-engineering systems.\u201d<\/p>\n

    \"Sequencing<\/a>In essence, the InSPECT framework suggests that the purposes of a digital object are determined by uses required by specific stakeholders. These purposes determine functionality and in turn, the properties that are needed over a period of time. By concentrating on these aspects, the suggestion is that an institution may develop a speedier, simpler and cheaper strategy for preserving resources over time.<\/p>\n

    In InSPECT, FBS \u2013 Function \u2013 Structure \u2013 Behaviour, becomes CCRSB:<\/p>\n