[Paper Review] Genericity versus expressivity - an exercise in semantic interoperable research information systems for Web Science

Christophe Guéret, Tamy Chambers+3 more|arXiv (Cornell University)|Apr 21, 2013

Semantic Web and Ontologies29 references4 citations

TL;DR

This paper proposes a middle-range, semantically interoperable research information system model that balances genericity and expressivity by combining a core international ontology with national and local extensions. It enables scalable, traceable data aggregation for science modeling, using Linked Data standards like FOAF, BIBO, and PROV-O, and demonstrates its feasibility through a VIVO-based implementation in the Netherlands.

ABSTRACT

The web does not only enable new forms of science, it also creates new possibilities to study science and new digital scholarship. This paper brings together multiple perspectives: from individual researchers seeking the best options to display their activities and market their skills on the academic job market; to academic institutions, national funding agencies, and countries needing to monitor the science system and account for public money spending. We also address the research interests aimed at better understanding the self-organising and complex nature of the science system through researcher tracing, the identification of the emergence of new fields, and knowledge discovery using large-data mining and non-linear dynamics. In particular this paper draws attention to the need for standardisation and data interoperability in the area of research information as an indispensable pre-condition for any science modelling. We discuss which levels of complexity are needed to provide a globally, interoperable, and expressive data infrastructure for research information. With possible dynamic science model applications in mind, we introduce the need for a "middle-range" level of complexity for data representation and propose a conceptual model for research data based on a core international ontology with national and local extensions.

Motivation & Objective

• Address the lack of consistent, interoperable data models for research information across institutions, nations, and research disciplines.
• Enable comprehensive modeling of science dynamics by integrating researcher-centric data beyond traditional bibliometric indicators.
• Support science modeling at multiple scales—from individual researchers to global science systems—by establishing a retraceable, aggregatable data layer.
• Overcome limitations of existing systems (e.g., author ambiguity, siloed data) through standardized, semantically rich representations.
• Facilitate dynamic science modeling by proposing a core ontology with extensible national and local layers for improved data interoperability and expressivity.

Proposed method

• Design a conceptual model based on a core international ontology, extended by national (e.g., Dutch NARCIS) and local (e.g., KNAW) ontologies.
• Leverage established Linked Data vocabularies such as FOAF (persons), BIBO (publications), LODE (events), SKOS (skills), and PROV-O (provenance) for semantic expressivity.
• Host the core ontology at the W3C, national extensions at national bodies (e.g., VSNU for the Netherlands), and local extensions at institutional repositories.
• Integrate the model with existing systems like VIVO and CERIF, using semantic mappings to ensure data reusability and interoperation.
• Apply the model to real-world data from the Dutch research information system (NARCIS) and a VIVO deployment to validate feasibility and scalability.
• Ensure extensibility and reusability by aligning with W3C standards and the JoinUp GLD initiative for governmental linked data.

Experimental results

Research questions

• RQ1 How can a research information system balance genericity for interoperability with expressivity for detailed modeling of researcher activities?
• RQ2 What level of data granularity enables both aggregation and deconstruction for science modeling across different scales of the science system?
• RQ3 How can national and local research data be semantically integrated with international standards to support large-scale science system analysis?
• RQ4 What role do semantic vocabularies (e.g., FOAF, BIBO, PROV-O) play in enabling traceable, interoperable, and extensible research information systems?
• RQ5 Can a core ontology with extensible layers support dynamic, non-linear science modeling while preserving data provenance and retraceability?

Key findings

• A middle-range data representation layer—neither too abstract nor too specific—enables both aggregation and deconstruction of research data for science modeling.
• The integration of core vocabularies (FOAF, BIBO, SKOS, LODE, PROV-O) with a modular ontology architecture supports high expressivity and semantic interoperability.
• Hosting the core ontology at the W3C, national extensions at national bodies (e.g., VSNU), and local extensions at institutions ensures scalable, trustworthy governance of the data stack.
• The proposed model successfully supports a VIVO-based implementation in the Netherlands, demonstrating feasibility in real-world deployment.
• The model enables the inclusion of non-traditional researcher indicators (e.g., career development factors) not covered by existing standards like RIS or CERIF.
• By aligning with W3C and JoinUp standards, the model ensures long-term sustainability and reusability across international research information systems.