Bulk Recovery Methodology For Legacy Assets
0 likes270 views
The document discusses developing a methodology to catalog and index eight years worth of learning objects in order to facilitate their finding, reuse, and sharing. It outlines challenges with traditional metadata description approaches and describes a preliminary methodology developed that involves: 1) Harvesting existing file properties and identifying duplicates for available metadata 2) Adding metadata specific to individual learning objects like title, description, and classification 3) Adding common metadata like copyright and pedagogy to groups of files 4) Converting the metadata to an XML format for transportation and validation The goal is to develop integrated workflows to standardize metadata capture during the learning object creation process.
![[Recovery! – Legacy Data] Start from the course environment Identify Learning Objects (What is a…) Find the files on Group file server Document 22 required metadata elements (Normetic, a LOM application profile) Apply best practices for accuracy, uniformity and quality](https://image.slidesharecdn.com/BulkRecoveryMethodologyForLegacyAssets10-123194590716-phpapp01/85/Bulk-Recovery-Methodology-For-Legacy-Assets-4-320.jpg)
Bulk Recovery Methodology For Legacy Assets
- 1. Robert Grégoire Business Development & Research
- 2. Our Goal A Librarian's Task Inadequacy of Traditional Metadata Description Methodologies Mass Management of the Data A Preliminary Methodology Results Reconciliation & Future Developments
- 3. Index eight years worth of learning objects in order to find, retrieve, reuse and share resources. Context: A multimedia development team Many projects Many actors for each project Life span of average project: 18 months Different uses and levels of aggregation Anticipated benefits: market GTA’s expertise, enhance productivity and reduce costs
- 4. [Recovery! – Legacy Data] Start from the course environment Identify Learning Objects (What is a…) Find the files on Group file server Document 22 required metadata elements (Normetic, a LOM application profile) Apply best practices for accuracy, uniformity and quality
- 5. File server does not support many to many relationships, thus forcing files duplication Finding files (naming conventions) and keeping track of described resources is a major problem Web interface and screen layout constrict work Repetitive nature of task conducive to errors Group edits impossible Metadata ownership denied, surrendering of learning objects sometimes forced Obligation to rely on editor’s code for conformance and interoperability
- 6. Harvest file properties for available metadata (size, date, format, etc.) - 2/22 Run Cyclic Redundancy Check to identify file duplicates Add resource-specific information: Title, Description, Key Words and Classification – 4/22 Add information common to groups of files, e.g.: Copyright, Contribution, Pedagogy, etc. – 16/22 (+ 2, GTA’s internal)
- 8. Step Actor Control Management 1. Automatic File Properties Harvesting Directory Report 30 Built in 1.1 General: Identifier (file name - ID) 2. Identify Duplicates Cyclic Redundancy Check Built in 3. Add MD specific to LO Librarian & Programmer Application developed: Modify tab 7.2.2 Relation: Resource.Description (file path) 9.2.2 Classification: TaxumPath.Taxum 4. Add MD common to groups of files Librarian & Programmer Queries will become application features 5. Convert MD to XML transportation format Programmer script LOM Validator 4.3 Technical: Location 3.1.2 Metametadata: Identifier.Entry
- 9. 1.1.2 General: Identifier.Entry (optional) File naming must be standardized yet recognizable: « Cardio_InsuffisanceMitrale.jpg » Unique file ID: concatenate classification + database ID + extension: « 612.1-107056.swf » 3.1.2 Metametadata: Identifier.Entry (optional) Is unique file ID with « .xml » extension
- 10. 4.3 Technical: Location (required) 128 bit encrypted (MD5) http://coral.ci.umoncton.ca/object_server/object.php?id=c15d8395f7cbd77e5065301f7c910e79 Object.php script resolves to learning object and allows flexibility to add other management features such as rights, permission, etc.
- 11. 7.2.2 Relation: Resource.Description (optional) Document 7.1 Relation: Type, e.g.: "is part of" 7.2.2 establishes file paths in the form of: Bank Course Module. E.g.: Consortium national de Formation en Santé | SINF 4255 - Soin complexe et critique | Examen clinique en soins critiques
- 12. 9.2.2 Classification: TaxumPath.Taxum (required) Dewey Decimal Classification, others to come Guides file server organization, file naming and systems
- 14. 1999 – 2008 Content Development Methodology development time to stable solution Workflows analysis to capture MD during development phase Reconcile server image to server reality in near future Feb. 2008: Server image for development purposes 2008 - Integrated MD Capture Methodology
- 15. Barriers to the emergence of LOs as a standard in the knowledge economy are identified Solutions for the recovery of legacy assets metadata are suggested The solution for an integrated MD documentation methodology involves workflows to standardize the capture of assets MD throughout the process of creation Ideally, only Title & Description (from context, e.g. course environment) need to be added before publishing asset to LOR