The multilingual semantic web has a featured role in the United Nation’s Food and Agriculture Organization. Work began with the Agrovoc Thesaurus, a multilingual, structured and controlled vocabulary designed to cover the terminology of all subject fields in agriculture, forestry, fisheries, food and related domains, and has progressed to the development of the Agrovoc Concept Server that is a first step towards an ontology service. The conversion process is being completed and a test version has been deployed, says FAO information and knowledge management officer Margherita Sini. When all is said and done, it will be the basis for developing specific domain ontologies that account for multilinguality and the localized representation of knowledge, and it can lead to additional and better services for users.

Stakeholders range from the FAO itself to students studying agricultural improvements in China or India to scientists who need information about specific organisms to policy makers who need various agricultural statistics to farmers in remote areas who may need to access services – such as information they don’t have today on new ways of dealing with crop pests, for instance—with the help of specialists at village centers that support Internet access. “The idea is to convert into a concept service or kind of ontology where we have concepts, terms and relationships and also can extract part of it and create domain-specific ontologies and build another application on top of it,” she says. “Generally on projects we are working to meet the needs of a variety of users so multilinguality is important. Not everybody speaks English.”

The Agrovoc thesaurus is semantically oriented in having relationships between the keywords it contains that are related to agriculture – for example, there is a generic relationship that a specific bug is a pest of rice plants. As this thesaurus completes its journey into a concept server, more information can be applied to specific types of relationships and you can start to gather meaning within and across concepts, Sini says. “In a thesaurus everything is a term but there are no clear distinctions between the terms that form a concept from terms that form another concept,” she says. With the Concept Server, you wind up with a collaborative reference platform and a “one-stop” shop for a pool of commonly used concepts related to agriculture, containing terms, definitions and relationships between terms in multiple languages derived from various sources.

“Multilinguality comes in because in a traditional thesaurus you had a one to one translation,” she explains. “But we know there are terms that may be translated with multiple terms in another language, or we know that for a preferred term in English the exact translation is not the same in Russian. That may be a synonym of the preferred term in English. We created the project to convert the thesaurus into a concept server where there are several concepts and we distinguish different terms in different languages, allow multiple relationships, allow for the fact that we have acronyms, synonyms, and near synonyms.”

Today in traditional searching systems someone types in terms like cow to find documents related to cows, she notes—generally in English and as it appears in the indexing system. “But now we can have any synonym, acronym or spelling variant—in any language—then we look in the ontology and see if the term is there,” she notes. “So a user might type ‘female dairy cattle’ and we find cows. And he can do it in any language, and we allow disambiguation.” URIs associated to every concept catalogue information, so no matter what language you are searching in you’re routed to the appropriate URI and get the documents that are associated to that URI.. For example, vessel could mean ship or blood vessel. “Using this kind of ontology we would look for the term vessel and in any concept it appears we would ask the user which way he meant the term vessel,” she says. And after he selects the one he means, it executes the search in the indexing system for that ID to have better and more precise search.

Says Sini, “We want to spread the concept of agricultural services based on semantics.”

Semantic Tech & Business Conference Returns to San Francisco

Semantic Tech & Business Conference returns to San Francisco in June! Join us from June 3-7 for complete coverage of Big Data, Linked Data, Extreme Information Management, and Semantic Web. From breakthrough approaches to solving business problems to the big data implications of fast–evolving technologies, SemTechBiz provides you with an unparalleled interactive experience and delivers tangible business value. We're offering a special early rate when you register by February 17. Sign up now!