Ontologies at FAO

AGROVOC

In order to convert the Agrovoc thesaurus in an Ontology, it was repurposed as an RDFS file in 2003. This allowed it to be visualized and modeled as an ontology using the KAON tools developed at the University of Karlsruhe.

Further improvements of the semantics of Agrovoc are being brought about through a project in collaboration with the Kasetsart University (in particular, with the AGRIS centre in Thailand). The aim of this project is to extend the traditional set of relations (BT, NT, RT, etc.) to include additional ones such as partOf, isA, hasPest, etc. to represent the semantics of the Thai plant domain. Furthermore, extensive work has been done to redesign the traditional thesaurus as an ontology using a relational database schema. A draft model of the database is currently under testing (available upon request to  FAO-Agris-Caris@fao.org.

In light of the possible benefits that semantic technologies can provide to knowledge management and information retrieval, not only is Agrovoc being converted to a concept based structure using the Ontology Web Language (OWL), but a portal will be designed that expands the services beyond the ones that are currently offered to users (e.g. information specialists, translators, developers, ontology modelers), extending the range of possible applications to which Agrovoc can be put.

Related Resources

Reengineering AGROVOC to Ontologies

From thesaurus to Ontology [ppt]

Demo file [KAON Format]

top

A project has been carried out in collaboration with other FAO departments to produce a domain ontology for food safety, animal health, and plant health. Starting from the AGROVOC Thesaurus and terms extracted from web sites and documents, a Food Safety and Animal Health Ontology has been created and modeled using the KAON tools.

This ontology has been used to develop an online portal. Some work still needs to be done with regard to ontology relationships.

Related Resources

Overview [ppt]

Demo [AVI]

Demo [GIF]. Note: To open the animated gif image, drag the gif into the browser window.

Download [KAON FORMAT (3 files)]

top

The objective of this project, carried out in collaboration with the Italian National Research Council (CNR), was to build a Fisheries Ontology by integrating the following fisheries and fisheries-related terminologies:

Aquatic Sciences and Fisheries Thesaurus (ASFA Thesaurus)

AGROVOC: Multilingual Agricultural Thesaurus (Agrovoc Thesaurus)

oneFish list of keywords

oneFish Topic headings/titles

FIGIS terms

Other sources were also used in the development of the ontology:

ASFIS Subject Categories and Scope Descriptions

Relevant subject glossaries

Relevant handbooks and scientific publications

Relevant websites

The ontology is now available in OWL files and an RDFS-KAON file. The work of subject matter experts is now needed in order to correct the structural problems that remain from the automated processing.

Related Resources

Overview [ppt]

Demo as Gif [Part 1], [Part 2]Note: To open the animated gif image, drag the gif into the browser window.

Demo as Avi [Part 1], [Part 2]

Download the ontology : [OWL Format], [Snapshot]

Download the ontology : [KAON Format], [Snapshot]

Publications on-line

[Sun SITE Central Europe (CEUR)] Proceedings of the EKAW*04 Workshop on Core Ontologies in Ontology Engineering - Northants. (UK), October 8, 2004.

For further information on ontology development (projects, conference presentations, journal articles, see the Publications page on this site.

An efficient and effective browsing system for the Food, Nutrition and Agriculture (FNA) journal has been developed under the auspices of the FAO Nutrition department. The system uses a simple knowledge model to guide the user's search and browse experience within FNA.

In order to achieve this result an ontology was developed as follows:

metadata of two bibliographical databases were combined;

RDFS was used to represent the semantic relationships between the metadata objects;

concepts, relations, attributes, and lexicalizations were accessed programmatically using the Kaon tools

The portal allows users to browse by keywords, categories, authors, etc. In addition, users can traverse resources and properties through the ontological relationships that connect them.

These steps are just a starting point for further exploitation of other semantic relationships.

Related Resources

Overview [ppt]

Demo [On-line]

Download [KAON Format]

top