Create get_bert_keywords.py

get_bert_keywords.py

@@ -0,0 +1,12 @@
+# import data manipulation libraries
+from nltk.corpus import stopwords
+from keybert import KeyBERT
+from collections import Counter
+# KeyBERT model
+keybert_model = KeyBERT(model='all-MiniLM-L6-v2')
+
+def get_bert_keywords(text, num_keywords, ignorewords):
+    result = keybert_model.extract_keywords(text, keyphrase_ngram_range=(
+        1, 1), top_n=num_keywords, stop_words=stopwords.words('english')+ignorewords)
+    keyberts = [keyword for keyword, score in result]
+    return keyberts