In the “A language learners’ forum” DangerDave2010 proposed a simple and efficient dictionary based lemmatizer
#encoding: utf8
lemmaDict = {}
with open('lemmatization-es.txt', 'rb') as f:
data = f.read().decode('utf8').replace(u'\r', u'').split(u'\n')
data = [a.split(u'\t') for a in data]
for a in data:
if len(a) >1:
lemmaDict[a[1]] = a[0]
def lemmatize(word):
return lemmaDict.get(word, word + u'*')
def test():
for a in [ u'salió', u'usuarios', u'abofeteéis', u'diferenciando', u'diferenciándola' ]:
print(lemmatize(a))
test()
The dictionary lemmatization-es.txt can be downloaded from lexiconista.
I downloaded some dictionaries to C:\Users\batagelj\Downloads\data\lemma .