Logo Lanfrica
  • Accueil
  • Atlas
  • Analyses
  • Documentation
  • Sign in

© 2026 Lanfrica. Tous droits réservés. Tous les droits d'auteur des ressources affichées sur le site Web Lanfrica appartiennent aux détenteurs de droits d'auteur d'origine, sauf indication contraire explicite.

EndoLink: A Knowledge Graph-Based Platform for Crowdsourced Endonym and Place Name Collection

Domaine:

natural language processinggeospatial

Type de record:

software
Créateur:
HinAhm
Éditeur:
UniUniHu,Mon
Éditeur:
The
Hôte:avatar
This paper introduces EndoLink, a web-based crowdsourcing platform that enables users to explore, collect, and contribute missing location names through a map-based interface. The platform addresses the under-representation of local languages in knowledge graphs, which limits the performance of natural language processing tools for geographic information extraction, particularly for low-resource languages. We present a quantitative analysis of place name coverage in Wikidata for Iran, Turkey, and Thailand, revealing a strong bias toward official and Western languages. User feedback indicates that EndoLink improves usability and lowers the barrier to contribution. Additionally, an evaluation with large language models demonstrates their limited ability to retrieve missing labels for endonyms in Central Kurdish, highlighting dependencies on existing multilingual data. EndoLink is publicly available at janinelaura-hindermann.gith….

Visit

doi.orgwww.zora.uzh.ch

Tags

410 Linguistics000 Computer science, knowledge & systems