Upload app.py

app.py

@@ -37,7 +37,7 @@ Hence, implementing named entity recognition before sentiment analysis helps to
 
 Using RNN-based approach can also overcome lexicon issues, but it also takes more resources.
 '''
-dat_name = 'E:/Project/NLP/news_sentiment_analytics/news_db/merged_news_data_' + datetime.today().strftime('%Y-%m-%d') + '.csv'
+dat_name = './news_db/merged_news_data_' + datetime.today().strftime('%Y-%m-%d') + '.csv'
 news = pd.read_csv(dat_name, on_bad_lines='skip')
 news[['pos_sent', 'neg_sent']] = news[['pos_sent', 'neg_sent']].fillna('')
 news['clean_content'] = news.clean_content.apply(lambda x: ast.literal_eval(x))