Hi, I have a difficulty between picking controlled vocabulary and a full blown knowledge graph.
in the context of semantic search that is dedicated at non programmers, does a full knowledge graph make a difference over a skos thesaurus? How do we make a typical end user benefit from it. Said differently, how to make the case of a full knowledge graph (ontology based) over a taxonomie to make the search smarter such that the typical end user can benefit from it? does someone has a specific example of how real entities (knowledge graph) modelization would out perform a thesaurus, in the context of end user experience? My guess is that it usefull for faceted browsing, as the user can exploit the related information. Otherwise, how, if the user do not specify them, can the semantic search engine exploit them. It could give the opportunity to exploit the properties of the entity or that related to an entity. But when it comes to assigning class, if an entity belong to several class that can be tricky. In thesaurus, you use compound names like Barack_Obama(senator). However for ontologies how would you do with class to find, information related to barack_Obama as a senator rather than as a president. The resource would have to be explicitly describe with the type of barack_oabama you are talking about ? I don't see the outperformance here. Many thanks -- Maatari Daniel Okouya Sent with Airmail
