Abstract

To integrate APIs from different domains and data providers, data need to have a well-defined meaning: imagine combining healthcare APIs using different units for human temperature, or financial APIs using different currencies.

This talk presents a semantic interoperability strategy based on:

• API (OAS3.0) and Schema (RDF) catalogues;
• specifications to attach RDF metadata to OAS3.0 API definitions and JSON Schema data models;
• tools to support API design and validation according to Italy's national interoperability guidelines.

The presentation covers ongoing standardization efforts (e.g., the registration of the OpenAPI and YAML-LD media types), a rationale of the specifications, and a demo of the Schema Editor: a tool providing a user-friendly interface to design semantic APIs.

Description

This talk will present strategies and python tools to create semantically interoperable REST APIs. After the problem statement, various solutions will be presented, including:

• contract-first api development with OpenAPI 3, ontologies and controlled vocabularies like the ones published by the European Union;
• the rdflib and pyld python libraries for processing json-ld and RDF files;
• the use of centralized catalogs such as schema.org.

Prerequisites:

• no prior knowledge of semantics and ontologies;
• practical experience with OpenAPI, json schema and data modeling and API design in general.

• an API catalogue describing APIs with OAS3.0;
• a schema registry with RDF ontologies and controlled vocabularies;
• a specification based on JSON-LD and JSON Schema to
  attach semantic metadata to (existing) API definitions and data models.