GraphDB 9.2 Supports RDF-Star to Match the Expressivity of Property Graphs

RDF-Star (also known as RDF*) enables the more efficient representation of scores, weights, temporal aspects and provenance in knowledge graphs

Sofia/Bulgaria Wednesday, April 29, 2020

Ontotext has released the 9.2 version of GraphDB – the most powerful RDF database for creating and maintaining knowledge graphs. In addition to the anticipated support for RDF-star (or RDF*)/SPARQL-star (or SPARQL*), it brings improvements in the plug-ins for semantic similarity and versioning. The new Proof plug-in provides inference explanation.

Knowledge graphs are the next generation tool for helping businesses make critical decisions based on harmonized knowledge models derived from siloed source systems. For this reason, more and more organizations are turning to knowledge graphs for better enterprise knowledge management, data and content analytics.

GraphDB 9.2 upgrades the modeling expressivity of the database with RDF*/SPARQL* – the most widely supported extension of those standards, which will be part of RDF 1.2. In RDF* one can make statements about statements, formally called statement-level annotations. For example, one can provide a time span for an employment relationship such as: (Joe works-for BigBucks) from 1/1/2018 to 30/3/2020. This mechanism goes beyond the expressivity of Property Graphs, where one can attach key-value pairs to relationships. The statement-level annotations enable a more efficient representation of scores, weights, temporal restrictions and provenance information.

Ontotext has invested substantial time and effort into developing comprehensive and efficient support for RDF* and SPARQL* in GraphDB. We have made extensions to GraphDB’s indices to make sure SPARQL* queries are handled in a fast and predictable manner. We also took care to handle properly specific cases like nested statement annotations and annotations on deleted statements. To guarantee backward and forward format compatibility, GraphDB introduces “embedded triple” (a reference to a triple) as a new RDF type. As a result, GraphDB reduces by more than 40% the number of RDF statements needed to model graphs with a complex metadata like Wikidata, the loading time and the required storage space. Ontotext has contributed the development effort for the RDF*/SPARQL* data format and corresponding changes to the parsers to the RDF4J open-source community.

In addition, the 9.2 release includes important plugin improvements. Now, the semantic similarity plugin, while refreshing the similarity indexes, remains online. This feature will support all production installations using graph or text embedding similarity indexes, so all indexes remain functional even during a rebuild.

Another substantial improvement is that now the data history and versioning plugin takes much less resource when comparing the current state with a previous period. GraphDB 9.2 features a new plugin to trace back all rules to generate a particular implicit statement.

The GraphDB team stays up-to-date with the latest known security warnings reported by all major vulnerabilities databases. The GraphDB 9.2 database and all integrated third party software are free from any publicly known security vulnerabilities.

Get the latest GraphDB now!

For more information, contact Doug Kimball, Chief Marketing Officer at Ontotext