Postponed: Towards Unified Approaches to Graph Models - RDF, Property Graph, SQL

POSTPONED because of Corona. A new date will be announced as soon as possible!

Welcome (13:00)
By Barbara Olsen, Danish Data Architects and Modelers and SAS Institute, and Christian Olsen, IDA.

Graph Standards, Communities and Challenges (20 minutes)
By Thomas Frisendal, Independent Graph Data Architect

The graph database universe is a combination of de facto and emerging standards. Thomas will do his best to give a quick overview of the “combatants”: (1) The openCypher and Apache Tinkerpop property graph communities, (2) Emerging standards: Property Graph extensions to SQL and the forthcoming GQL (Graph Query Language) by the ISO IEC/JTC1/SC32/WG3 committee (the “SQL committee”) backed up by the LDBC graph language task forces as well as the US INCITS ad hoc groups, (3) The existing W3C (Worldwide Web Consortium) and its’ semantic standards such as RDF and OWL and, finally, (4) The RDF* and SPARQL* proposals by Olaf Hartig.

Use Case: Building Semantic Data Hub with LPG, GraphQL and Kafka Streams (30 minutes)
By Per Jensen and Artur Mrozowski, Tryg Forsikring

In this talk, we’d like to present some hands on implementations of ideas flourishing within graph technologies space. At Tryg we use property graph data modelling and apply it to ACORD, an insurance industry standard data model based on RDF. Property graph approach offers powerful opportunity to design GraphQL schemas. We try to apply Domain Driven Design principles to the data platform and move towards domain oriented data ownership. The domains are exposed as number of GraphQL APIs endpoints offering easy access to non-semantically trained end users as well, as resource-oriented approach . All of this is framed by Event Driven Architecture where we use both Kafka Streams and GraphQL subscritptions to push messages to the consumers.

The RDF*/SPARQL* Approach to Graph Database Interoperability incl. GraphQL and Graph Databases (45 minutes)
By Olaf Hartig, Linköping University

In this talk I will introduce the RDF*/SPARQL* approach to capture and to query statement-level metadata (edge properties) in the RDF context, and I will show how this approach provides a foundation for achieving interoperability between RDF and Property Graphs. As a bonus, I will present some of our research results related to GraphQL, including an adaptation of the GraphQL schema definition language to define schemas for Property Graphs.

A Tale of two graphs: The Property Graph and the RDF Graph (45 minutes)
By Jesus Barassa, Neo4j

In this talk, I propose a comparative analysis of the two main approaches to implementing Knowledge Graphs: LPG and RDF. Where do they come from and how are they similar and different. I'll also explore a number of ways in which the two approaches (LPG and RDF) can coexist and complement each other, like: Integration/Interoperability or Knowledge Management/Inferencing. I'll try to include a little demonstration of some of these topics in action on Neo4j.

Wrapping up (16:15 for 10 minutes)
By Thomas Frisendal, Danish Data Architects and Modelers

Event organizers

The event is co-arranged by these 2 meetups: Danish Data Architects and Modelers (this one) and Data Talks (IDA).

Speaker list: https://drive.google.com/open?id=1mFst_LQI-m2IvnKXHi1j6txUYN5Wl64g