Giants join forces to improve agricultural semantics


I am writing this post on my way back to home from Rome where I participated in the Improving Agricultural Semantics workshop that was held in Food and Agricultural Organisation of United Nations and which was funded by Bill and Melinda Gates foundation.


Arriving I decided not to speak about the situation in Greece and get as much as possible from this event where all the important organisations of the agricultural sector will be present. Of course you understand that this was difficult because everybody was asking me about the referendum. I broke my promise 🙁

The workshop started with a very interesting session about the results of the pre-workshop survey about the Knowledge Organisations Systems and data used and produced by organisations and the main problems that they are addressing. The survey has very interesting insights and you can find it here. In addition, the methodology and results of creating GACS (Global Agricultural Concept Scheme) were presented.

A series of interesting presentations about use cases on solving real problems using semantic technologies, metadata standards and linked data followed. These use cases triggered discussion about how important is standardisation and mapping between standard vocabularies in the digital age where data of different types are generated in big volumes every day. The diversity of data and problems that the data producers, data consumers and technology providers are facing helped the participants to define the main challenges that should be addressed during the next years. These are

  • to define specific use cases of using semantics to improve the agricultural information access, sharing and discovery
  • to define the required infrastructure and tools required for sharing and discovery of agricultural and nutrition information
  • to define the principles for improving sharing, access and discovery of agricultural and nutrition information

The second day was devoted to break out sessions that aimed at addressing each of the above challenges. More specifically, three different groups were formed one for each challenged. Every group worked on the challenge and reported the results back to the plenary at the end of the day.

Although we are discussing and working on agricultural semantics for many years we all agree that the complexity and diversity of data is such that we need:

  • to align the efforts on defining standards that the community can adopt. Not to define just another standard but to identify the gaps and define the one that are missing
  • to develop open tools and guidelines in how to use them and make them available and searchable through a global registry. CIARD RING could play that role but it should be enhanced by including guidelines and community based functionalities that could allow the exchange of best practices on improving agriculture information sharing and access
  • to develop capacity building activities such as trainings and educational material about open data, standards and open tools that can help people with no expertise in linked open data to share effectively their information

Global initiatives like GODAN and organisations like Open Data Institute could help in promoting and communicating the principles, the standards, the tools, the infrastructure and the good practices.

As a major conclusion, one could say that the whole community need to join forces to improve the agricultural semantics. Existing giants such as The Food and Agriculture Organization of the United Nations (FAO), the CAB International (CABI), and the National Agricultural Library of the USA (NAL) should work together with all organisations of the agriculture and nutrition sector to facilitate the information sharing, access and use. Small players’ contribution is also very important to that end.

Giant thesauri such as the AGROVOC Concept Scheme, the CAB Thesaurus (CABT), and NAL Thesaurus (NALT) should be linked and made available for anyone in the agricultural and nutrition sector that wants to connect his data to the web of data. The linking should be done in a seamless way for the data producers and consumers not necessarily forcing them to annotate their content with another vocabulary. We should be democratic in the approach: leave them to have their own vocabulary that fits to their needs and align this vocabulary to the giants.

Leave a Comment.

Can you prove that you are not a bot? * Time limit is exhausted. Please reload CAPTCHA.