tokenize_texts = [ ['mentioned', 'reviewers', **'episode', 'exactly'**] ]
porter_stemmed_texts = []
for i in range(0, len(tokenize_texts )):
porter_stemmed_text = [nltk.stem.PorterStemmer().stem(word) for word in tokenize_texts[i]]
porter_stemmed_texts.append(porter_stemmed_text)
porter_stemmed_texts
output :
[ ['mention', 'review', **'episod', 'exactli'**] ]
expect output :-
[ ['mention', 'review', **'episode', 'exactly'**] ]
Are these errors normal. Can't we get 100% accurate words.
The stemmer is working as intended.
"Episode" should stem to "episod" so that it stems the same way as "episodic".
"Exactly" -> "Exactli" is an a quirk in the algorithm, but it doesn't make a difference in the end because you should also be stemming the text you're comparing against, so it will also contain 'exactli' once stemmed.