- From: John S. Erickson <john.erickson@hp.com>
- Date: Wed, 7 May 2003 14:57:37 -0400
- To: "'SIMILE public list'" <www-rdf-dspace@w3.org>
Mick writes: > Rather than risk severe scope creep for the work > underway, I suggest that we view these problems as > a basis for Use Case(s) based upon the History System. > Perhaps we should update the History System Use Case > in the SIMILE Research Drivers document [SRD] to > reflect this? This is a great point. The "Drivers" document anticipates that the History System will enable the following: "...To be trusted, such a steward must be able to retain and provide non-repudiatable audit trails of how content and metadata has been altered, migrated, transformed, or augmented over time (``Proof of transformation path``, or ``data provenance``). The steward may need to undo and/or redo part of the transformation path based on errors discovered, improved tools, etc. Queries on the information regarding transformation paths across the corpus can be expected in the process of optimally maintaining and preserving the corpus..." This requires a richness of property declaration/collection surrounding curatorial events that (I don't believe) is currently captured in Jason's document. Yes, all of the collection-level stuff is in there, but there needs to be models for a set of activities AT A HIGHER LEVEL than the current repository-level, pushing-bits-around actions. I realize that the statement of work says that "...(t)he intent is not to extend or enhance the amount of information included in the output from the History System, nor to create a more complex model..." but essentially to make the information it produces more usable. But I fail to see how the History system can do its job documenting "...events of archival interest (that) occur within the system (for example, when a community is created, an item's instance metadata is edited, or the members of a collection are modified)..." if a richer halo of metadata around those events is not created. That is a hypothesis. It can be tested by examining use cases that demonstrate collection of metadata surrounding curatorial activities, and esp. subsequent use of that metadata to guide e.g. rollback, etc. I'm esp. interested in retrieving explicit information that helps humans inthe second case understand e.g. WHY something was done, rather than than the implied meaning of e.g. transformations, migrations, etc. John
Received on Wednesday, 7 May 2003 14:59:15 UTC