1. Introduction

Here we introduce the semantic framework document's audience, definitions, and MMI approaches to fulfilling it.

Document Audience

This document is written for managers, programmers, and semantic web researchers dealing with knowledge representation and the mediation of environmental data, in particular marine data.

We intend to show managers how these technologies can address their needs for exchanging semantic information. Believing that the technologies represent a viable approach, we describe how adoption of this approach is likely to further completion of a working solution. We also review potential concerns about initial adoption of a specific approach to semantic interoperability.

We ask the computer science community to closely evaluate the proposed architecture and offer feedback. The following information is especially welcome:

  • What, if anything, in this overall architecture is not coherent or viable?
  • What system components are missing or presented poorly?
  • What is the best way to achieve progress overall?
  • What other efforts can help move this effort forward, and how can this effort move other efforts forward?
Semantic framework links Data Provider and Data User

Definitions and Concepts

Additional definitions are provided within later sections.

Semantic framework

The 'semantic framework' described by this document is an architectural design concept addressing knowledge management and concept mediation across multiple systems and components. It is a conceptual artifact, rather than a concrete set of objects.

Infrastructure (components)

The infrastructure, or a set of infrastructure components, are the elements of a system that are typically not visible to the end user, but are necessary for other elements to interoperate. This term refers to the actual elements of the designed or implemented system.


A Term is a string of characters representing a concept; typically (but not always) the string of characters has a human-recognizable meaning associated with the represented concept.


A URI is a Uniform Resource Identifier, a concept defined by the World Wide Web Consortium. A URI is a web name for a specific resource; the resource itself may or may not be accessible via the web. A URI may be either a URL (Uniform Resource Locator), like you enter in a web browser, or a URN (Uniform Resource Name), a unique string that describes a resource. The primary intent of a URI is to provide a unique identifier for terms. The URI is the basic unit of information in this semantic mediation framework.

Controlled Vocabulary and Ontology

A collection of terms is called a controlled vocabulary. A controlled vocabulary is as a set of restricted words, used by an information community when describing resources or discovering data. A controlled vocabulary prevents misspellings and avoids the use of arbitrary, duplicative, or confusing words that cause inconsistencies when cataloging data.

An ontology is "a set of representational primitives with which to model a domain of knowledge or discourse" (Tom Gruber, http://tomgruber.org/writing/ontology-definition-2007.htm, to appear in Encyclopedia of Database Systems. For this document, an ontology can be considered a formalized expression of a controlled vocabulary.

A controlled vocabulary can be published in an ontology in plain RDF (Resource Description Framework), SKOS (Simple Knowledge Organization System), or OWL (Web Ontology Language). RDF defines relationships between terms in a relatively simple way, using subject-predicate-object expressions. SKOS and OWL are based on RDF, and define additional types of resources that can be used to describe or define relationships.

The preferred choice for a controlled vocabulary among these ontology formats depends on the application.

MMI Involvement

The tools and services required for the initial operational framework can be developed by a small team. Most of them are planned for development by November, 2008 for a vocabulary mapping workshop hosted by the Marine Metadata Interoperability project. If more people contribute to the tools, the framework (and demonstration applications) will provide a greater level of functionality, and value added, for the Data Originators and data users.