Framework and Registry for Brokering and Mediation Components

04 Feb 2016

Framework and Registry for Brokering and Mediation Components

Revisions

See revised Case Statement V2 and TAB review response in attached documents.

 

 

 

Proposal for ICSU-WDS/RDA WG on a Framework and Registry for Brokering and  Mediation Components

Background
Mediation and Brokering

In a SOA framework, software components interoperability is implemented by defining and using common protocols. In a web services framework, interoperability protocols  are characterized by their interface methods, and bindings as well as their payload content. For example, in the case of protocols for data discovery and access, the payload content contains data and metadata encoded using specific models.   

Mediation and adaptation modules are often used to map two different content models or two different interface methods or two different binding types. Commonly, a mediation module addresses the mediation of one feature characterizing two different protocols -e.g. the payload content model.  

Brokering services can be used to implement more advanced and general mediation functionalities. Brokering components address all the three protocol heterogeneities: methods, bindings, and payload content models. In addition, they implement mediation from many-to-many different protocols  

In the geospatial information domain,  Brokering services were successfully introduced to mediate across the different disciplines/Community protocols for data discovery, access, transformation and (in the most advanced cases) processing.  

  

Problem Description
The scope of the WG covers data sources from research and scholarly communication. Within this domain obstacles for wider application of brokering techniques are:

1.    Multiple service protocols for data discovery, access, and application or processing;
2.    Multiple content standards for data and metadata, augmented by Community profiles and non-standard implementations;
3.    Multiple vocabularies and ontologies.
4.    Multiple adaptation and mediation modules that are not guaranteed to be compatible.

Research projects and research data infrastructure initiatives often solve problems associated with this diversity as a matter of course, but the knowledge gained and components developed during such a process are not visible and useful to others. Furthermore, project life cycle limitations lead to lack of sustainability, loss of expertise, code, and infrastructure.

 

Address the Need
●    Define a description schema for services, vocabularies, ontologies, content standards, and adaptation components that allow services and clients to be matched – with a mediation component interposed if required.

●    Establish a prototype registry based on the above.

●    Describe a collection of existing mediation and adaptation components that can interoperate through well-defined existing interface specifications and applicable standards.

●    Create a mediation and adaptation components registry -the objective is to support implementation of a more general and agnostic mediation capability.

●    Define a test bed environment for testing interoperability of mediation alternatives leading to recommendations for application areas. The focus will be on metadata and data mediation across data systems that address different disciplines and scopes. 

 

Mediation functions to be supported by the test bed:

1.    Data discovery and access protocols including harvesting and synchronous distribution (subscription/notification).
2.    Data and metadata content transformation (harmonisation).
3.    Metadata content enhancement and Linked Open Data enablement through vocabularies and ontologies.
4.    Application to popular protocols and service definitions.


Read the full case statement