First Commit

app.py

@@ -0,0 +1,86 @@
+import streamlit as st
+import datetime
+import pandas as pd
+from pygooglenews import GoogleNews
+from transformers import pipeline
+import plotly.graph_objects as go
+
+# Load the sentiment analysis model
+pipe = pipeline("text-classification", model="pramudyalyza/bert-indonesian-finetuned-news")
+
+# Function to process the keyword and get sentiment analysis
+def process_keyword(keyword):
+    one_week_ago = datetime.datetime.now() - datetime.timedelta(days=7)
+    gn = GoogleNews(lang='id', country='ID')
+    search = gn.search(keyword)
+
+    filtered_headlines = []
+    for entry in search['entries']:
+        published_date = datetime.datetime.strptime(entry['published'], '%a, %d %b %Y %H:%M:%S %Z')
+        if published_date > one_week_ago:
+            filtered_headlines.append(entry['title'])
+
+    df = pd.DataFrame(filtered_headlines, columns=['title'])
+    df_clean = df.drop_duplicates()
+
+    df_clean['sentiment'] = df_clean['title'].apply(lambda x: pipe(x)[0]['label'])
+
+    positive_count = (df_clean['sentiment'] == 'Positive').sum()
+    negative_count = (df_clean['sentiment'] == 'Negative').sum()
+    total_count = len(df_clean)
+
+    return positive_count, negative_count, total_count, df_clean
+
+# Streamlit app layout
+st.title("News Sentiment Analysis Dashboard")
+
+keyword_input = st.text_input("Enter a keyword to search for news", placeholder="Type a keyword...")
+
+if st.button("Analyze"):
+    if keyword_input:
+        positive_count, negative_count, total_count, df_clean = process_keyword(keyword_input)
+
+        # Create plots
+        fig_positive = go.Figure(go.Indicator(
+            mode="gauge+number",
+            value=positive_count,
+            title={'text': "Positive Sentiment"},
+            gauge={'axis': {'range': [0, total_count]},
+                   'bar': {'color': "green"}}
+        ))
+
+        fig_negative = go.Figure(go.Indicator(
+            mode="gauge+number",
+            value=negative_count,
+            title={'text': "Negative Sentiment"},
+            gauge={'axis': {'range': [0, total_count]},
+                   'bar': {'color': "red"}}
+        ))
+
+        fig_donut = go.Figure(go.Pie(
+            labels=['Positive', 'Negative'],
+            values=[positive_count, negative_count],
+            hole=0.5,
+            marker=dict(colors=['green', 'red'])
+        ))
+        fig_donut.update_layout(title_text='Sentiment Distribution')
+
+        # Display results
+        st.plotly_chart(fig_positive)
+        st.plotly_chart(fig_negative)
+        st.plotly_chart(fig_donut)
+        st.write(f"News articles found: {total_count}")
+
+        # Show DataFrame
+        st.dataframe(df_clean)
+
+        # Download CSV
+        csv = df_clean.to_csv(index=False).encode('utf-8')
+        st.download_button(
+            label="Download CSV",
+            data=csv,
+            file_name='news_sentiment_analysis.csv',
+            mime='text/csv',
+        )
+    else:
+        st.error("Please enter a keyword.")

requirements.txt

@@ -0,0 +1,5 @@
+pandas==2.2.3
+plotly==5.24.1
+pygooglenews==0.1.2
+streamlit==1.37.1
+transformers==4.36.2