Metadata and Metamodeling


Metadata Models

Annotea. In Annotea, annotations are modeled as RDF metadata classes and created by document authors by choosing an annotation class or type from a list of possible options. One can inherit from an existing annotation type a new type defined in a user RDF Schema, and add it into the list for future use. Annotations comprise a body, a link, the author’s ID and additional properties. The body forms the content of annotations which may be either included in the RDF document or linked as an external document. The link is a URI of the document (or resource) annotated. The author’s identification and other properties are RDF properties which may contain third party metadata models such as Dublin Core and user defined attribute-value pairs.

RDFSuite. Different forms of serializing RDF descriptions in XML poses a problem in evaluation of RDF semantics. Even though they represent a single RDF graph, RDF description syntaxes in XML vary. Systems that process RDF descriptions must form a RDF graph that represents the semantics of the descriptions so that the validation can be performed for all forms of RDF serializations. To solve this problem, RDFSuite introduces a formal data model for RDF description bases that forms a complete set of validation constraints as well as an expanded type set for the RDF Schema. The validator in the RDFSuite system utilizes the such constraints and types to form a RDF graph for the evaluation of RDF semantics.

SAM. SAM is based on the Slide project [7, 98], theWebDAV [63] implementation of the Apache open-source group. WebDAV, an extended version of the Hypertext Transfer Protocol (HTTP), allows Web users not only to browse online documents, but also to modify, remove, or add new content onto the Web site, forming a basis for distributed document authoring. However, more importantly, Slide provides a single programming and access interface for resources as well as associated metadata to manage. One can use a single URL to access a resource and its metadata with two different methods on the same server. Many WebDAV client applications are also commonly available, making the WebDAV servers such as Slide attractive to distributed systems developers.

SAM and CMCS do not suggest a common metadata model to developers or users, however, internally, all the metadata is stored in attribute-value pairs with specialized property semantics. Slide associates each resource a URI and several predefined and user-defined properties. Properties can be in XML format. SAM names properties so that semantic relationships can be formed between resources. For example, a resource property, named hastranslations, can hold a list of translations that can be applied on the resource while another property, named children, can point a list of other resources that carry a parental relationship.

IKM uses the Rights Markup Extensions [61] to the Extensible Rights Markup Language (XrML) [30] to model the rights of indigenous groups on multimedia content. The model is based on an analysis of tribal laws across indigenous communities, revealing several common factors that decides access to traditional knowledge, such as native/non-native restrictions, a users membership to a particular clan, status, role, gender, and so on.