pierreguillou
commited on
Commit
·
57f5b08
1
Parent(s):
c6ab8b3
Update app.py
Browse files
app.py
CHANGED
@@ -32,13 +32,13 @@ kw_model = {
|
|
32 |
pos_pattern='<CONJ.*>*<ADP.*>*<ADV.*>*<NUM.*>*<ADJ.*>*<N.*>+'
|
33 |
|
34 |
# download stop words in Portuguese
|
35 |
-
import nltk
|
36 |
-
nltk.download('stopwords')
|
37 |
-
from nltk.corpus import stopwords
|
38 |
-
stop_words = set(stopwords.words('portuguese'))
|
39 |
|
40 |
# define o vectorizer
|
41 |
-
vectorizer = KeyphraseCountVectorizer(spacy_pipeline='pt_core_news_lg', pos_pattern=pos_pattern, stop_words=
|
42 |
|
43 |
# function principal (keywords)
|
44 |
def get_kw_html(model_id, doc, top_n, diversity):
|
|
|
32 |
pos_pattern='<CONJ.*>*<ADP.*>*<ADV.*>*<NUM.*>*<ADJ.*>*<N.*>+'
|
33 |
|
34 |
# download stop words in Portuguese
|
35 |
+
#import nltk
|
36 |
+
#nltk.download('stopwords')
|
37 |
+
#from nltk.corpus import stopwords
|
38 |
+
#stop_words = set(stopwords.words('portuguese'))
|
39 |
|
40 |
# define o vectorizer
|
41 |
+
vectorizer = KeyphraseCountVectorizer(spacy_pipeline='pt_core_news_lg', pos_pattern=pos_pattern, stop_words=None, lowercase=False)
|
42 |
|
43 |
# function principal (keywords)
|
44 |
def get_kw_html(model_id, doc, top_n, diversity):
|