[Paper Review] Semantic Integration & Single-Site Opening of Multiple Governmental Data Sources.

Konstantinos Kotis, Iraklis I. Athanasakis+1 more|arXiv (Cornell University)|Jul 2, 2014

Semantic Web and Ontologies14 references3 citations

TL;DR

This paper proposes S3-AI, a semantic application integration approach that enables unified, ontology-mediated querying of multiple distributed governmental data sources from a single site using D2RQ and Fuseki. The method preserves data ownership and autonomy while supporting real-time, federated SPARQL queries over live, evolving data, demonstrating effectiveness in an e-Government IT helpdesk use case with minimal cost and high customizability.

ABSTRACT

In many cases, government data is still "locked" in several "data silos", even within the boundaries of a single (inter-)national public organization with disparate and distributed organizational units and departments spread across multiple sites. Opening data and enabling its unified querying from a single site in an efficient and effective way is a semantic application integration and open government data challenge. This paper describes how NARA is using Semantic Web technology to implement an application integration approach within the boundaries of its organization via opening and querying multiple governmental data sources from a single site. The generic approach proposed, namely S3-AI, provides support to answering unified, ontology-mediated, federated queries to data produced and exploited by disparate applications, while these are being located in different organizational sites. S3-AI preserves ownership, autonomy and independency of applications and data. The paper extensively demonstrates S3-AI, using the D2RQ and Fuseki technologies, for addressing the needs of a governmental "IT helpdesk support" case.

Motivation & Objective

• To address the challenge of integrating disparate, distributed governmental data sources within a single public administration organization.
• To enable unified, real-time querying of live, evolving data from multiple autonomous data sources without centralizing or redesigning existing systems.
• To support semantic application integration using open, lightweight Semantic Web technologies while preserving data ownership and autonomy.
• To demonstrate the feasibility and benefits of the approach in a real-world e-Government IT helpdesk use case.
• To identify best practices and lessons learned for deploying semantic technologies in public sector data integration.

Proposed method

• Using D2RQ to map relational databases to virtual RDF graphs, enabling semantic access to existing relational data sources.
• Employing Fuseki as a SPARQL 1.1 endpoint to execute federated queries across multiple virtual RDF sources.
• Developing a domain-specific reference ontology and using it to mediate queries across heterogeneous data schemas.
• Creating semantic mappings between application-specific schemas and the reference ontology using a semi-automated approach with manual validation.
• Implementing a single-site access point that exposes a unified SPARQL endpoint for federated queries over distributed data.
• Using dummy data for public demonstration to ensure data security and privacy during initial deployment.

Experimental results

Research questions

• RQ1 How can multiple autonomous, distributed governmental data sources be semantically integrated without centralizing or modifying their original systems?
• RQ2 To what extent can D2RQ and Fuseki enable efficient, real-time, federated SPARQL querying over live, evolving data in a government setting?
• RQ3 What are the practical challenges and best practices in mapping heterogeneous data schemas to a common reference ontology in a public administration context?
• RQ4 How can data ownership and autonomy be preserved while enabling unified data access and querying?
• RQ5 What are the performance and maintainability trade-offs of using lightweight, open-source Semantic Web technologies for government data integration?

Key findings

• The S3-AI approach successfully enabled unified, ontology-mediated, federated SPARQL querying across five distributed data sources within NARA’s e-Government IT helpdesk system.
• The use of D2RQ and Fuseki allowed real-time access to live, evolving data without data replication or system redesign.
• The approach demonstrated low maintenance effort, high customizability, and low cost, making it suitable for deployment in public administrations with limited resources.
• The integration process was feasible even with complex schema heterogeneity, especially when using a well-chosen reference ontology and semi-automated mapping support.
• The deployment revealed challenges in manual mapping, including incorrect term pairings, syntax errors, and missing mappings, highlighting the need for careful validation.
• Data security remains a critical concern, with the authors noting that only anonymized or dummy data was exposed publicly, and secure endpoints are a priority for future work.