The UN Food and Agriculture Organization (FAO) has a number of (online) publications (e.g., “Food, Nutrition and Agriculture”). As all such sites, they also have to offer an easy way to search through the articles, find the right papers based on topic, keywords, concepts, etc. The approach is to add an interface to a local search engine that would also offer such concept-based search. Potentially, such a search engine could also span over several related datasets that share the same concepts. To achieve this, one of the issues to solve to have the concepts themselves expressed in a machine readable form.
An organization like the FAO has an accumulated knowledge that goes back to several decades in the form of, for example, thesauri or glossaries. The upcoming W3C SKOS specification provides a tool to encode such thesauri, glossaries, terminologies in a standard, machine readable format. By publishing it in a standard format, not only is the thesaurus usable for FAO’s own purposes but it can also be reused by anybody else on the Web. This is what the FAO has done by publishing their Agrovoc Thesaurus in SKOS, and base their search interface on that thesaurus, too. Similarly, other institutions like libraries can publish their catalogues, subject headings, glossaries, etc, using the same format, providing a web of interrelated concepts and terms that could be used to organize and integrate data overall. This can be an important step in the advancement of the (Semantic) Web.
For further details on the FAO application, please look at the Semantic Web Case Study published on the W3C site (courtesy of Gauri Salokhe, Margherita Sini, and Johannes Keizer). And look at the SKOS specification (or the SKOS Primer) if you want to know more technical details!