Finding Ontologies

Ontologies are generally developed around communities, and thus you may already be familiar with some that are being used in your area of interest. However, besides asking within your community, there are several techniques you can use to find ontologies that may be relevant to your needs.

Ontology registries and libraries are excellent ways to discover and share ontologies within a community of users. Semantic-oriented search engines often create indexes of ontologies, even providing ways to work with the ontologies in their index, and general Web search engines like Google index ontologies that can be searched.

General Search

The least sophisticated way of locating ontologies is to use existing Web search technology, like Google, and limit the search results to a particular filetype. In Google, you would do this with the following query:

temperature filetype:owl

This will return results that contain the word temperature, limiting the result set to OWL files.

This method has its limitations, such as Google only recognizes files with extensions. This means that a hosted OWL file without the extension (gcmd-science instead of gcmd-science.owl) will not appear in Google’s results.

Registries and Libraries

Ontology registries are generally domain specific, meaning that a community working on similar issues or with similar data will collaborate to provide a service for their community. In their most advanced form, registries can enable interoperability by storing relationships between ontologies, allowing them to act as mediators between different software tools that rely on the relationships between ontologies.

Libraries are generally straight indexes of ontologies, categorized using a set of metadata to provide a catalog of ontologies. However, there are more advanced libraries that allow searching using the semantic query language, SPARQL. Both libraries and registries offer opportunities to find ontologies that may be of use in your work.

The following are registries provided by various ontological communities. Some are applicable to the marine science community, others are provided as examples of how registries function.

MMI Ontology Registry and Repository

MMI has deployed the MMI Ontology Registry and Repository, a version of the BioPortal ontology application, which allows users to access and share ontologies in use in the marine science community. While targeted initially at marine concepts, the MMI Ontology and Registry is open to ontologies from all environmental science fields.

BioPortal

BioPortal is a Web application developed by the National Center for Biomedical Ontologies. It provides the biomedical community a space in which to share and discover ontologies and is the foundation of MMI’s Ontology Registry and Repository.

National Cancer Institute Bioportal

The National Cancer Institute (NCI) BioPortal is an older version of the NCBO’s BioPortal software. It provides access to the NCI Thesaurus, the NCI Metathesaurus, and other select and publicly accessible biomedical terminologies hosted at the NCI.

OntoSelect Ontology Library

The OntoSelect Library provides a catalog of ontologies ranging from cars to plants. There are search and browse functions provided for users and the library tracks metadata for ontologies, including format, language, and domain.

DAML Ontology Library

The DARPA Agent Markup Language Library contains records of ontologies from around the Web. Information from this source may be outdated.

Protégé Ontology Library

The Protégé Library is a flat list of user-provided ontologies hosted on the Protégé project wiki. No metadata is tracked and no search functionality is provided, though the entire list can be searched in a browser.

Search Engines

Unlike Google, which is a general index of the web, there are tools available which allow you to search indexes of ontologies. The power and flexibility of these tools vary, from simple interfaces like Google to more complex products that allow you to run queries in SPARQL.

Swoogle

Swoogle, as the name implies, is an attempt to replicate the usefulness of Google in the semantic web realm. It allows searching for ontologies, documents, and terms, and has an index of over 10,000 ontologies.

ONTOSEARCH2

ONTOSEARCH2 allows users to search on keyword or use SPARQL for more advanced queries and is under active development. It is currently unknown how many ontologies it includes in its index.

Falcons

Falcons describes itself as, “a keyword-based search engine for the Semantic Web, equipped with browsing capability. Falcons provides keyword-based search for URIs identifying objects, concepts (classes and properties), and documents on the Semantic Web. Falcons also provides a summary for each entity (object, class, property), integrated from all over the Semantic Web.”

Watson

Watson is a keyword search engine for semantic documents that provides great information regarding classes, properties, and individuals in ontologies.

Other Semantic Search Engines

The number of semantic search engines is increasing. You can find a current list, according to the w3c Linking Open Data Task Force and the w3c ESW group.

Suggested Citation

Alexander, P. 2011. "Finding Ontologies." In The MMI Guides: Navigating the World of Marine Metadata. http://marinemetadata.org/guides/vocabs/ont/existing/finding. Accessed December 11, 2019.