Linked Linguistic Knowledge Graph

Resource

The Linked Linguistic Knowledge Graph (LLKG) is a language-independent linguistic resource designed for diachronic analysis set on well-known ontologies for linguistics. LLKG is suitable for holding information from various collections by interconnecting the entities with external resources to expand the reachable knowledge. 

The resources involved in this study are the Linguistic Knowledge Graph, a time-sensitive graph database for linguistic knowledge, and Etymological WordNet, a lexical resource for describing word origins. Except for the significance of the data they hold, these resources were considered aligned with the objectives of this work due to their design choices, which hinder the contribution to the cloud of Linguistic Linked Open Data and prevent the discovery of new knowledge. Figuratively speaking, we intend to burst the bubble of their isolation. To this purpose, this work focuses on translating the Labelled Property structure into Resource Description Framework Schema and adopting the lexicon model for ontologies LEMON. This work also illustrates how to enrich the graph by manually linking its entities to resources on the Web, e.g., Universal WordNet, LiLa and Wikidata.

Citation

Please cite

@article{ghizzota2025enhancing,
  title={Enhancing Linguistic Resources for Diachronic Analysis via Linked Data},
  author={Ghizzota, Eleonora and Basile, Pierpaolo and d’Amato, Claudia and Fanizzi, Nicola},
  journal={GWC 2025},
  pages={192},
  year={2025}
}

Release

The code for generating the resource is available on GitHub.