Semantic Search is the technology search engines employ to better understand the context of a search. Is not just about simply discovering information any more, it’s about ordering the web into a coherent flow of information; how topics, themes, ideas, text, video, audio is connected and related to each other.
A lot of data on the Web has been published in a form that that makes it easy for us humans to digest. Just think of weather reports, flight schedules or business news. Semantic Web aims to make this information also understandable to computers so that they can act on that information.
Acting on the information must rely on a non-ambiguous description of content and its relationship to other pieces of information.
Resource Description Framework (RDF)
RDF is a data model which the World Wide Web Consortium (W3C) defined in 1999. The data model is simple yet powerful enough to express properties of Web based objects or to express properties of anything which can have a URI address. RDF uses triples which identify a resource, a property associated with the resource and the value of this property. The value can in turn be another resource, which allows RDF descriptions to encode simple and complex relationships depending on the domain.
RDF alone would work if everybody used same URIs for the same things and the same properties. Now that this is not the case, reconciliation of RDF data models from multiple sources can be realized with the help of ontologies and rules in general. An ontology is a formal description of concepts and properties used in semantic metadata descriptions but not only that – Ontologies provide us with primitives that help us discover implicit relationships.
JSON-LD extends JSON to the Linked Data domain where the data structures expressed in JSON have a context which is uniquely identified with International Resource Identifiers (IRI). This approach allows JSON-LD make descriptions about unique objects and refer to other objects which was not available in plain JSON.
Those who have worked with Resource Description Framework (RDF) data model are also happy to hear that JSON-LD can express the RDF data model but the users of JSON-LD need not have prior experience with RDF in order to utilize the data.
WEB ONTOLOGY LANGUAGE (OWL)
OWL provides proven primitives to generate more RDF from your RDF. In other words you can create more value from your existing information assets with the help of an OWL compatible Rule Engine.
For realizing the Semantic Web vision and enabling search engines to provide more accurate search results, the W3C has also specified RDF a technology for embedding RDF descriptions inside HTML documents. This allows companies to become more reachable via search engines and with a limited cost when their Web Content Management software can embed rich metadata inside the previously presentation-rich and semantically-poor pages.
SPARQL query language
SPARQL comes from SPARQL Protocol and RDF Query Language. Similar to the role SQL has played in the evolution of relational algebra and its implementation in relational databases, SPARQL provides a common access interface for repositories that are either native RDF databases or have established a SPARQL endpoint on a non-RDF database.
The W3C has specified SPARQL in three parts: