Skip to Navigation

NCBO User Profile: Jin-Dong Kim, Database Center for Life Science

Submitted by rcasuga on June 27, 2012 - 12:48

 OntoFinder/OntoFactory: ontology production based on BioPortal

spacer

"Ontology engineering can be greatly synergized by the spirit of sharing"

Jin-Dong Kim, Ph.D
Project Associate Professor
Database Center for Life Science

Research Interests: Life science is a resource-rich area due to high throughput experimental technologies and its nature of publicness. There are many groups working for the integration and organization of the rich resources of life science, and Database Center for Life Science (DBCLS) is one of them. Particularly, DBCLS emphasizes the semantic aspect of database integration, pursuing active application of Semantic Web (SW) technology. Ontology development is in the heart of database integration project of DBCLS, enabling semantic connection of data to be integrated.

Use of NCBO services:

Ontology development effort in DBCLS is based on the NCBO's BioPortal service, which we are recognizing as the central repository of ontologies. Figure 1 illustrates a scenario of ontology production based on BioPortal, which we have figured out to be ideal.

Ontology selection
With BioPortal, a user, who seek to have an ontology for representation or organization of his/her own data, can search the repository to find relavant ontologies, which we call base ontologies.
Term selection
Among them, only relevant parts may be collected with unnecessary parts filtered-out. We call the collection a seed ontology.
Ontology development
If necessry, new terms may be created and added to the seed ontology, making it to better fit to the target data (step 3).
Ontology registration
When the user feels ready, the developed ontology may be registered back to BioPortal (step 4), to keep the cycle going.
spacer

Figure 1. Ontology production based on BioPortal

To encourage and support the ontology production based on BioPortal, DBCLS is developing the OntoFinder/OntoFactory system, aiming at providing non-computer experts with an easy interface to assist the ontology selection and term selection steps, until a user produces his/her seed ontology based on BioPortal.

 

OntoFinder/OntoFactory is a web service for seed ontology development based on BioPortal. It begins from ontology search on submission of multiple terms. The search result is rendered in a table as shown in Figure 2. In the table, the row headers (antibodies and TLR4 in the example) correspond to the input terms, and the column headers (IMRGO, and so on) correspond to the ontologies which are the results of search. The individual ontology terms that match the input terms are shown in the corresponding cells. In this way, the table view shows the ontology search results at both ontology and term levels at the same time

 

In the table view, ontology columns are draggable and foldable so that a user can change the order of ontologies or hide unnecessary ones. In this way, OntoFactory support the ontology selection step of ontology production. The individual terms are also draggable, and a user can collect terms of hie/her interest into the column,selected terms. The collected terms then can be exported to the local storage as a OWL (Web Ontology Language) file. In this way, OntoFactory supports the term selection step until the user get his/her own seed ontology.

 

We believe ontology engineering can be greatly synersized by the spirit of sharing, and deeply grateful to BioPortal for providing the repository of ontologies as the central place of sharing. We hope the OntoFinder/OntoFactory service to contribute to the effort of ontology sharing through BioPortal, thus contributing to the integration of life science resources also.

spacer

Figure 2. A screenshot of OntoFinder/OntoFactory service

  • Login or register to post comments
gipoco.com is neither affiliated with the authors of this page nor responsible for its contents. This is a safe-cache copy of the original web site.