Constructing a Dutch Historical Climate Vocabulary (v1.1) based on GLOBALISE

This version represents an intermediate, exploratory stage of development and is not intended for direct application without further manual annotation and refinement. Subsequent work (v1.2+) focuses on manual curation and formalization of the vocabulary. v1.1 of the Dutch Historical Climate Vocabulary builds on v1 by identifying candidate terms for enrichment. A pre-trained GLOBALISE Word2Vec model (van Wissen, 2024; DOI 10.5281/zenodo.15038313) was evaluated for semantic coherence, and a FastText model was trained to capture additional candidate terms for inclusion. Candidate terms from the FastText Model were incorporated into v1.1 of the vocabulary based on similarity thresholds and other statistical metrics. This version is exploratory and requires further annotation and validation to be operational. This version was initially shared internally on 2026-01-22 and is deposited here for provenance purposes.