Search

This paper shows the experience of using AGROVOC thesaurus as a basis and model of information and knowledge organization and domain representation applied to soybean and sugar cane agricultural intensification in Brazil. A textual corpus of about 2,5 million of words was compiled and candidate terms extracted automatically for creating an initial hierarchical conceptual map, then compared to AGROVOC terms. The results show the importance of AGROVOC as a resource to organize and represent agricultural information and knowledge, as a reference for creating new terminological products on agriculture, and open new possibilities for enrichment of AGROVOC´s Brazilian Portuguese terminologies.

The Linked Open Data (LOD) initiative is the first large-scale attempt at realizing the Semantic Web vision of Tim Berners-Lee. As of September 2011, there are almost 300 knowledge bases linked together and available for public access through both web browsers as well as semantic applications. However, the task of building the links between knowledge basesis still a laborious manual task and as the LOD cloud continues to grow rapidly,the task becomes more daunting. This paper looks at the AGROVOC thesaurus developed and maintained by the United Nations Food and Agriculture Organization (UN FAO), and how a framework can be developed to semiautomatically discover links between AGROVOC and other knowledge bases on the LOD.