AGROVOC data can be accessed in different ways depending on the user's needs. If a user wants to browse AGROVOC, search for concepts or terms, or look at the structure, the web browsing interface is the best option. If the user wants to use content of AGROVOC, extract parts of it, or embed it in applications, then the best choice is to download it or to use one of the programmable interfaces, like the SPARQL endpoint, the REST API or the SOAP web services.
For simple web/based browsing, AGROVOC can use the Skosmos Search & browse interface. It is an open source web-based SKOS browser and publishing tool. This interface offers search and browse functionalities, alphabetical and thematic index, structured concept display, visualized concept hierarchy and multilingual user interface. The version of AGROVOC loaded in Skosmos is always the latest release of AGROVOC LOD.
Web-based use of the SPARQL interface
There are other web-based interactive ways to use AGROVOC, like the web-based SPARQL interface which requires specific technical knowledge and the results are in machine-readable formats. There are a number of potentially useful things that can not be done with the browsing interface, e.g. download part of the tree or view or download a full sub vocabulary. Additionally, it is not possible to choose which properties to see in the tree for each concept.
On the other hand, the SPARQL endpoint has a web-based interface where results can be displayed in tables and can be downloaded as a CSV. Such use of the SPARQL interface may be handy if the end user is not a programmer but needs something more than just browsing AGROVOC.
The SPARQL interface for AGROVOC is at http://agrovoc.uniroma2.it/sparql.
The difference between the Web-based uses and this one is the focus on services. These services are meant to be used from within software applications, and to return data in machine-readable formats. These formats are primarily RDF native formats like Turtle or N3 and RDF-enabled formats like RDF/XML or JSON-LD. RDF sets are not meant for human readability, but for machines. Typically, RDF data are either loaded into a triple store to be then reused or loaded into an RDF editor/viewer or parsed in an application, by developers who know how to handle RDF data