<!> These are my raw notes taken on my localized MoinMoin? wiki on the eXe laptop, then transferred here. Having your own local wiki is the only way to fly at these things, plus you look really cutting edge as well ;-) I'll try to condense at some point into a set of observations about what I learnt about metadata and what impact this day's discussions had on my thinking about metatada and the eXe.
Notes on the Educause Higher Education Metadata Workshops
Higher Education in a Global Context
John Mason
LOM-DCMI-IMS (http://metadatafaq...) didn't quite get this url, sorry.
http://www.cetis.ac.uk/metadatafaq ? -- DJM
Wine Bottle metaphor: content=wine, container=bottle, metadata=label
metadata frameworks
METS - metadata exchange transmission standard (http://www.loc.gov/standards/mets/)
SC36 http://mdlet.jtc1sc36.org/ - next generation metadata standard; first part of standard needs to be a framework. leveraging off of LOM and Dublin Core approaches. LOM group has not commenced any work on existing standard. standardization is an ongoing process:
- aimed at achieving consensus
- matching stakeholder reqs to solutions
- requirind maintenance, revision & updates
- drives innovation
more elements added (audience, accessibility)
DC-Kernel WG http://dublincore.org/groups/kernel/
proposed terms: who, what where when
IEEE LTSC
LOM data model
XML Binding (in ballot)
RDF Binding (mid 2005)
Accessibility Metadata
- defined in terms of abilities (matching resources to needs &/or preferences)
- IMS AccessForALl
- Dublin Core (dc:Accessibility)
W3C
- semantic web (RDF, OWL)
SCORM
- a market winner (widespread global adoption, esp. corporate training)
- reusable learning content - 'Units of Learning' (SCOs & SCAs; profiles LOM & adds mandatory info)
- not really a standard, collection of guidelines
- 3 main SCORM books (CAM, RTE, Sequencing & Navigation)
- an implementation model
Challenges
- accomodating user requirements
- developing stakeholder appreciation for role & scope
- maximizing benefits for adopting existing standards
- achieving and maintianing interoperability
- working with an agreed framework
- working with higher education stakeholders
http://reusable learning.org/
Towards a New Zealand Education Sector Architecture Framework
Mark Horgan
ESAF Educational Sector Architectural Framework
projects:
- data modelling, roles and responsibilities (ESDM)
- metadata schema (ESMA)
- integration services (ESIS)
- authentication / authorisation (ESAA)
- federated search (ESFS)
Data Roles and Models
custodianship of data? who is responsible?
Metadata Schema
uses Dublin Core - DC-ed
single authentication system
working with MAMS system for authentication
Metadata Research Outputs
Debbie Campbell
new types of annotation:
- provenance
- annotation (forms a type of peer review; see: NSDL in USA)
- accessibility
- co-locating - ability to find related information in easy manner. JISC: linking research.
peer review is current "standard" for academic research.
OAI protocol - metadata harvesting?
how will universities create metadata around research outputs?
generating metadata:
- citations
- e-portfolios
- automatic generation of metadata
Generation Software:
- improvement (JHOVE)
- recombination -
ARROW discovery service; national resource discovery across the australian higher education sector.
Metadata and Standards Development
Maxine Brodie - Macquarie University Library
what's the business case?
why have standards?
- safety
- quality
- compatability
- cost reduction
- technology transfer
eliminating avoidable costs:
- information seeking
- interpretation/decision making
- integration/re-organization
- information reporting
what is the context? ie. Library as context ...
- provides clues to controlled vocabulary, structures, syntax and rules.
- economies of scale
principles of metadata creation:
- control
- scale - assumes large quantity of 'data'.
- relevence
- heterogeneity
- homogeneity
- re-use
increasing view on semantic web that metadata is collaborative exercise.
metautopia or metacrap:
- people lie
- people are lazy
- people are stupid
- mission impossible - know thyself?
- schemas aren't neutral
- metrics influence result
- there's more than one way to describe something
Doctorow, 2001
Metadata in Practice: Defining an Education Sector Metadata Schema for New Zealand
Adrienne Kebbel & Mark Horgen
National Digital Strategy is also establishing a national elearning infrastructure.
centralised repository of national and cultural resources
finding, using, administering digital resources
XML - RDF -> DUBLIN CORE [DC-Lib DC-Ed DC-Gov]
NB: Dublin Core has organisational specificity(?) or is it just some kind of namespacing? DC-Ed - look into this.
Metadata Schema Deliverables
- XML schema for ESMS
- ESMS shared vocabulary
TeRG - Tertiary Elearning Reference Group
Metadata and Institutional Repositories: Managing Multiple Schemata
Geoff Payne - Monash University
ARROW project (http://arrow.edu.au/) - vehicle for an institutional repository in universities.
4 partners: monash, univ of new south wales, Swinburne Uni of Technology, National Library of Australia
What is an Institutional Repository?
- managed collection of digital objects
- institutional in scope
- consistent data and metadata
- enable discovery by communities of practice for whom objects are of interest
- read, input, export objects to facilitate sharing
- respect access constraints
- sustainable over time
- facilitating preservation
Technical Issues:
- Dublin Core - insufficiently granular
- LOM - not good for bibliographic metadata
- needed to preserve metadata already existing
Strategy:
- support metadata schemata to suit individual data models
- no requirement for all metadata to be shoehorned into a single schema
Two paths to interoperable metadata Godby, Smith and Childress. 2003.
Open Source Repository Software
http://www.soros.org/openaccess/software/
Fedora Consortium
Metadata in the ASPR Project
Kevin Bradley
Australian Partnership for Sustainable Repositories
3 projects:
- MAMS
- ARROW
- ASPR
NB: again Fedora project comes up.
principles of sustainability:
- for the user
- enduring
- accessible
- commitment by the institution.
Types of Material:
- research output
- digitised content
- "Born Digital" material (vital to long term research)
Sustainable Formats:
- image and sound (BWF .wav, TIFF)
- text (XML)
Metadata decisions:
- mandate certain fields in MODS
- D-Space and extending Dublin Core
- adhere to acknowledged standards
Sustainability:
- sustainability of raw data
- sustainability of access to meaning
- economics of sustainability
- economics of participation
- value of data
- tools, software and sustainability
metadata is critical to sustainability. preservation metadata
dissemination of knowledge and appropriate software is the aim of the project.
Case Studies
Universitas 21 Project
Tony Koppi
http://www.lrc3.unsw.edu.au:8040/
managers commissioned the development of software to:
- identify unpublished learning and teaching resources within U21
- enable them to be shared and reused
- make efficiency gains
- limiting reinventing the wheel
- raise quality by reusing the best available
Learning Resource Catalogue, now open to all HE Institutions
| Learning Tasks |
| Learning Resources | Learning Supports |
Description of LOs based on IMS metadata.
Submitting LO to LRC3:
Required Fields | Secondary Fields | Technical | Educational | Visibility
Sharing learning designs:
- no reward, unlike research
- awareness of learning objects and reusability
- feel obliged to create own course
- nothing as good as own creation
- saving time actually means more work
Swinburne's LILO
Derek Whitehead - Swinburne University of Technology
simplified version of LOM. Driven by needs of Content management system. To support digital objects used in CMS/LMS.
people require very simple interface to adding metadata. granularity issues of metadata applied to learning objects. manual application of metadata is impractical for ephemeral Learning Objects (?)
Experiences in the Management of Content at Griffith University. Metadata in Practice: Developing Metadata for the University's Digital Collections
Anne Loria - Griffith University
using Harvest Road Hive
looked at existing metadata schemas.
CanCore?: a subset of LOM
implement full metadata schema but use only elements required (via interfaces). trouble fitting some wanted metadata into existing schemas. Copyright requirements - major impact across Australia.
International LOM survey report - http://dlist.sir.arizona.edu/archive/00000403/
- underutilized elements
- innapropriate use
control vocabularies by using drop-down lists.
VCARD:
- suggested by LOM for contributor details
- vcard format though is generally ignored
- not XML encoded
- new standard? (FOAF?)
Thesauri & Taxonomies:
- images
- medical
- rights
- statement
- ownership
- description
- date of expiry
Art images, customized VRA-Core Schema
Using Hive to manage ePrints. OAI-PMH (Open Archives Initiative Protocol for Metadata Harvesting)
Learning Objects:
- determine metadata requirements
- LOM
- Feedback from stakeholders
Principles:
- shouldn't use metadata elements for other purposes
- who will create metadata?
- describe resource, not intended usage.
- data entry: efficient/intuitive