Update scrap_word_count_image.py

nicecoding1 · web-flow · commit 7944538a53de · 2021-06-11T17:52:03.000+09:00
diff --git a/scrap_word_count_image.py b/scrap_word_count_image.py
@@ -6,25 +6,35 @@
 from collections import Counter
 from wordcloud import WordCloud
 import matplotlib.pyplot as plt
-
+import nltk
+from nltk.corpus import stopwords
+import re
+from collections import Counter
 
 # 스크래핑
 url = "https://healthguides.cnn.com/finding-the-right-migraine-treatment/how-to-ensure-your-doctor-understands-your-migraine-severity"
 webpage = requests.get(url) 
 soup = BeautifulSoup(webpage.content, "html.parser")
 item = soup.select_one("article.ArticlePage-mainContent")
-data = item.text
+data = str(item.text).lower()
 
 # 단어 분리
-word = str(data).lower().strip().split()
-count = Counter(word).most_common()
-words = dict(count)
-# print(words)
+nltk.download('all')
+cleaned_data = re.sub(r'[^\.\?\!\w\d\s]','',data)
+words = nltk.word_tokenize(cleaned_data)
+tokens_pos = nltk.pos_tag(words)
+
+NN_words = []
+for word, pos in tokens_pos:
+    if 'NN' in pos:
+        NN_words.append(word)
+c = Counter(NN_words)
 
 # 워드클라우드
 wordcloud = WordCloud(font_path='C:\\Windows\\Fonts\\Gothic.ttf',
-                      background_color="white", max_font_size=100).generate_from_frequencies(words)
+                      background_color="white", max_font_size=100).generate_from_frequencies(c)
 
 plt.imshow(wordcloud, interpolation='bilinear')
 plt.axis('off')
 plt.show()
+wordcloud.to_file("wordcloud.png")
