Metadata
Content must be described appropriately so that users may discover it easily. The descriptive information associated with each repository resource (think library online catalog) is known as metadata. There are internationally recognized metadata standards to choose from such as Dublin Core which has 15 fields to the IEEE Learning Object Metadata(LOM) which when fully implemented has over 130 fields. All of the Dublin Core metadata fields are included in the LOM metadata schema. We recommend, at a bare minimum, that the Dublin Core fields be used. Sufficient metadata is the key to precisely locating the appropriate resource within your own repository. Metadata also enables the cross sharing of resources across repositories. For example, if Repository A uses Dublin Core metadata fields and Repository B uses IEEE LOM v 1.0, both repositories can communicate metadata to each other through a crosswalk of the related fields.
Dublin Core and the IEEE LOM both require common fields such as the resource title, author, description, relevant keywords, and copyright statements. Fields are classified as “required” or “optional”. A chart comparing these two commonly used metadata schemas (as well as those used by The Orange Grove as well as the repository of the University System of Georgia) is included in here [PDF]. Other elements may be required depending on which schema is being used. Because collections can quickly become extremely large, a key repository function is to perform targeted searches for resources. Associating robust metadata with each and every repository resource makes discoverability more likely, and makes it possible for users to locate resources that exactly match their specified needs.
The Metadata Team should review Dublin Core, LOM, or CanCore to select a metadata schema for the repository. You may wish to customize one of these schemas by adding additional fields or controlled vocabularies. In Florida, the postsecondary Statewide Common Course Numbering system database, the K12 Sunshine State Standards database, GEM Subjects and Assessment vocabulary, and the ERIC Thesaurus are controlled vocabularies embedded in the contribution wizard which generates the metadata for each resource. A controlled vocabulary is is an established list of standardized terminology for use in indexing and retrieval of information. An example of a controlled vocabulary is subject headings used to describe library resources. A controlled vocabulary ensures that a subject will be described using the same preferred term each time it is indexed and this will make it easier to find all information about a specific topic during the search process. (OECD Glossary of Statistical Terms, http://stats.oecd.org/glossary/detail.asp?ID=6260)
The Orange Grove began metadata entry by using the required fields recommended by the SREB-SCORE metadata guide. The SREB-SCORE group reviewed the IEEE LOM v 1.0 metadata and recommend requiring a sub-set of those elements that were most relevant to education in 2005. See SCORE User Guide. However, The Orange Grove has since reduced the number of required fields to approximately 19. See (provide chart of LOM fields with The University System of Georgia also began with the recommended SREB-SCORE LOM elements but has also reduced the number of required fields to approximately 15. A chart illustrating the various metadata fields can be found here.
It has been The Orange Grove’s experience that metadata presentation to the user is dynamic in that you may continually discover enhancements to improve system usability and metadata accuracy. For example, adding controlled vocabulary databases (e.g., statewide course numbers and ERIC Thesaurus terms) allows selection of terms via drop down lists, check boxes, and similar mechanisms. This speeds metadata information entry and improves its accuracy. The Orange Grove has incorporated the ERIC Thesaurus, Florida statewide course numbering system, GEM Subjects, GEM Assessments, and Florida K12 Sunshine State Standards as databases into its LOM v. 1.0 Classification field. Keep in mind that controlled vocabularies require periodic updating, which requires importing the updated database into the system.
It is also important to decide how to collect the metadata for each item. Will the user complete all or some of the metadata fields during contribution, or will a staff member add appropriate metadata to the record? We suggest having the contributor complete metadata with a metadata review metadata as part of your repository quality standards, at least during the initial implementation. The types of metadata errors discovered during the review will enable you to accurately target future training efforts or support needs. You should also consider whether the required metadata fields will vary based on the type of resource contributed. For example, an asset consisting of a single image will have less contextual information than a complete learning object, and you may wish to designate fewer fields as “required”.
The Orange Grove requires users to complete the “required” metadata fields when contributing an item to the repository. During the contribution process, metadata information is collected via a series of dropdown menus, radio buttons, text fields, and free text for a few fields. There are both optional and required fields. Contributors may elect to share ownership of an item to allow others to complete some of the metadata fields. The Orange Grove has a part-time metadata reviewer on staff who reviews the submitted metadata for accuracy, and individual institutions may also implement a review process to ensure accuracy of metadata.
Resources
