We offer a public service called URIBurner (a live instance of our Virtuoso Platform and its powerful Sponger Middleware Module) that leverages the power of a variety of NLP services, Entity Relationship Type Semantics, and LOD Cloud lookups for Entity Linking and Reconciliation.
Here’s a Knowledge Graph generation example, using a Blog Post that makes reference to Will Smith.
The entire extract, transform, and load process is packed into a series of actions associated with a single clickable URL (https://linkeddata.uriburner.com/about/html/https/towardsdatascience.com/the-building-a-large-scale-accurate-and-fresh-knowledge-graph-71ebd912210e).
Here are the different Knowledge Graph access and interaction documents generated as a result of the aforementioned URL lookup (or de-reference):
- About This Post — note DBpedia Identifier for Will Smith as one of the skos:related property values
- About This Post via a different Entity Description Document — i.e., one the includes Faceted Browsing whereby filtering by Entity Type (in addition to other Attributes) provides deeper follow-your-nose exploration
- SPARQL Query (also usable via SQL) — where the solution document (result set) includes information from other data sources and is obtainable in a variety of formats (e.g., CSV and JSON).
