WATERWHEEL (WATERloo Water and Hydrologic Entity Extractor and Linker) is a spaCy pipeline component that detects rivers, lakes, and other hydrologic entities, and links these entities to Wikidata. Although an acronym, we typically style the name as WaterWheel to avoid the impression that we're screaming all the time.
PyPI: https://pypi.org/project/waterwheel/
WaterWheel is written as a spaCy module. Sample usage:
import spacy
from waterwheel import WaterWheel
nlp = spacy.load('en_core_web_sm')
nlp.add_pipe(WaterWheel(nlp))
doc = nlp('The ultimate source of the Mackenzie River is Thutade Lake.')
for ent in doc.ents:
print(f'{ent.text}, type: {ent.label_}, {ent._.wikilink} [{ent.start_char}, {ent.end_char}]')
# Results:
# Mackenzie River, type: RIVER, https://www.wikidata.org/wiki/Q3411 [27, 42]
# Thutade Lake, type: LAKE, https://www.wikidata.org/wiki/Q7333634 [46, 58]