Ontologies power the Semantic Web
Ontologies come from information science and artificial intelligence. They are structures for modeling a domain of knowledge and that model is a key source for providing the context and meaning required to make the web semantic.
However, there is no single “right” way to model a domain – all ontologies have had some scope and framework to guide their development. This in part has led to a number of ontology standards which reflect the strengths and needs of different applications.
Linked Data accepts this, allowing multiple ontologies to be referenced to inform a particular RDF.
OWL vs. SKOS vs. zThes
Some elements, such as the ability to describe synonyms, or any type of equivalence relationship, associative relationships and add term information like descriptions and scope notes were available in early zThes definitions but not in SKOS. Subsequent extensions to the standards mean that there is now much overlap.
In 2007, Smartlogic choose to adopt the zThes standard as the application of Semaphore is very much based around the use of language and human interactions. As we can transform zThes to SKOS the internal standard used is largely irrelevant.
SKOS is built as a sub-set of OWL, however SKOS and OWL are intended for different purposes:
- OWL allows the explicit modelling/description of a domain.
- SKOS provides vocabulary and navigational structure.
Managing ontologies in Semaphore Ontology Manager
Semaphore Ontology Manager has utilities to convert SKOS to zThes and vice versa and provides a comprehensive tool for building this type of vocabulary and navigation structure.
It is not intended to be a full-blown OWL ontology management tool (like, for example Protégé), so the granularity and complexity of modeling is restricted. However the trade-off for complexity is ease of use – Ontology Manager can be used by Business Analysts or Subject Matter Experts to build up suitable vocabulary structures – using workflow and logic from a long, well developed discipline of library sciences to ensure the correctness and integrity of the final structure. Protégé is arguably a more specialist tool.
Semaphore provides the productivity workbench that allows non-specialist users to develop the ontology:
- Intuitive interface using a language familiar to information managers
- Text Miner to analyze text and extract out entities and noun phrases, automatically discovering topics and evidence in content.
- Term Copy to bring in sections of existing models for re-use or re-purposing
- Drag terms from spreadsheets, websites, etc. and contextual drop generates the appropriate relationship or term type.
- Review Tool to publish the model for review and collect and manage feedback comments from peers, subject matter experts, users, etc.
|
|
|