Presentation Open Access

The Labeling System: A bottom-up approach for enriched vocabularies in the humanities

Florian Thiery; Thomas Engel

Shared thesauri of concepts are increasingly used in the process of data modelling and annotating resources in the Semantic Web. This growing family of linked data thesauri follows a top-down principle. Vocabularies and broader concepts (SKOS-) are being created, maintained and provided under the supervision of central authorities to provide general and generic approaches used by scientists in the humanities. But the diversity of research questions in the humanities makes it virtually impossible to create shared controlled vocabularies that cover a wide range of potential applications, and therefore satisfy the needs of diverse stakeholders. Reliable interconnections among independent systems could solve this conceptual bottleneck of controlled vocabularies. The Labeling System (LS), developed by i3mainz and IEG in contrast follows a bottom-up approach, enabling scientists working in the digital humanities to manage, create and publish their own controlled vocabularies as a SKOS concept scheme and concepts provided via a REST API and URIs. One term of the vocabulary can be linked to broader corresponding concepts of domain experts and will become labels. The labels embed those broader concepts persistently into existing structures using a clean and straightforward UI. Technically the LS is defined over a flat ontology and can be queried through its triple store. The created concepts can then be interlinked with well known LOD resources from e.g. The Getty Research Institute or the British Museum, but also to authorities maintaining linked data resources from natural science domains. The LS is domain independent, while uniting perspectives of different scientific disciplines on the same label and therefore contributing to interdisciplinary collaboration for building up cross and inter-domain linked data communities. As the newly created expert resources are available persistently, the concept is quotable, which strengthens the scientific discourse of their semantic shape. The paper addresses principles of the Labeling System in the light of heterogeneous archaeological data from Western Europe and the Middle East. Consequently, “usual” archaeological topics of conceptualizing and interlinking temporal and spatial concepts (meaning) will be discussed. To what extent is it possible to align existing concepts with “inserting” specific concepts of domain experts? How can the LS be used to solve the ambiguity of a place type and its role or function in a specific archaeological meaning? Furthermore, we will show how the non-technical researcher can use the Labeling System to get introduced into the process of linked-data conceptualization. Finally, the paper details the benefits of enriching linked-data concepts through relating to linked-data communities of other domains, e.g. geology or anatomy.

Files (2.8 MB)
Name Size
2.8 MB Download
All versions This version
Views 7272
Downloads 3738
Data volume 102.3 MB105.1 MB
Unique views 6767
Unique downloads 3435


Cite as