Federating Scholarly Infrastructures with GraphQL

TL;DR

This paper introduces a GraphQL-based federated query service that enables integrated retrieval and interoperability of scholarly content across multiple distributed infrastructures, enhancing findability and reusability.

Contribution

It presents a novel federated query system for heterogeneous scholarly infrastructures and methods for linking artefact metadata and content using persistent identifiers and semantic resources.

Findings

Enables integrated querying across ORKG, DataCite, GeoNames

Improves interoperability through DOI-based linking and semantic resource connections

Enhances findability and reusability of scholarly knowledge

Abstract

A plethora of scholarly knowledge is being published on distributed scholarly infrastructures. Querying a single infrastructure is no longer sufficient for researchers to satisfy information needs. We present a GraphQL-based federated query service for executing distributed queries on numerous, heterogeneous scholarly infrastructures (currently, ORKG, DataCite and GeoNames), thus enabling the integrated retrieval of scholarly content from these infrastructures. Furthermore, we present the methods that enable cross-walks between artefact metadata and artefact content across scholarly infrastructures, specifically DOI-based persistent identification of ORKG artefacts (e.g., ORKG comparisons) and linking ORKG content to third-party semantic resources (e.g., taxonomies, thesauri, ontologies). This type of linking increases interoperability, facilitates the reuse of scholarly knowledge, and enables finding machine actionable scholarly knowledge published by ORKG in global scholarly infrastructures. In summary, we suggest applying the established linked data principles to scholarly knowledge to improve its findability, interoperability, and ultimately reusability, i.e., improve scholarly knowledge FAIR-ness.