Create app.py

app.py

@@ -0,0 +1,298 @@
+import streamlit as st
+import pandas as pd
+import numpy as np
+import re
+import nltk
+from nltk.corpus import stopwords
+from nltk.stem.porter import PorterStemmer
+from nltk.tokenize import word_tokenize
+from sklearn.feature_extraction.text import CountVectorizer
+from sklearn.preprocessing import LabelEncoder
+from sklearn.naive_bayes import MultinomialNB, BernoulliNB
+from sklearn.metrics import accuracy_score
+import pickle
+import os
+
+# Download NLTK data
+@st.cache_resource
+def download_nltk_data():
+    nltk.download('stopwords', quiet=True)
+    nltk.download('punkt', quiet=True)
+
+download_nltk_data()
+
+# Set page config
+st.set_page_config(
+    page_title="IMDB Sentiment Analysis",
+    page_icon="🎬",
+    layout="wide",
+    initial_sidebar_state="expanded"
+)
+
+# Custom CSS
+st.markdown("""
+<style>
+    .main-header {
+        font-size: 2.5rem;
+        font-weight: bold;
+        color: #1f77b4;
+        text-align: center;
+        margin-bottom: 1rem;
+    }
+    .sub-header {
+        font-size: 1.2rem;
+        color: #666;
+        text-align: center;
+        margin-bottom: 2rem;
+    }
+    .prediction-box {
+        padding: 20px;
+        border-radius: 10px;
+        margin: 20px 0;
+        text-align: center;
+    }
+    .positive {
+        background-color: #d4edda;
+        color: #155724;
+        border: 1px solid #c3e6cb;
+    }
+    .negative {
+        background-color: #f8d7da;
+        color: #721c24;
+        border: 1px solid #f5c6cb;
+    }
+</style>
+""", unsafe_allow_html=True)
+
+# Title and description
+st.markdown('<p class="main-header">🎬 IMDB Movie Review Sentiment Analysis</p>', unsafe_allow_html=True)
+st.markdown('<p class="sub-header">Powered by Naive Bayes | Built with Streamlit</p>', unsafe_allow_html=True)
+
+# Sidebar
+with st.sidebar:
+    st.header("⚙️ Settings")
+    model_choice = st.selectbox(
+        "Select Model",
+        ["Multinomial Naive Bayes", "Bernoulli Naive Bayes"],
+        help="Choose which Naive Bayes algorithm to use"
+    )
+    
+    st.markdown("---")
+    st.header("📊 Model Info")
+    st.info("""
+    - **Algorithm**: Naive Bayes
+    - **Features**: Bag of Words
+    - **Preprocessing**: 
+      - HTML tag removal
+      - Lowercasing
+      - Special character removal
+      - Stopword removal
+      - Stemming
+    """)
+
+# Preprocessing functions
+def remove_html_tags(review):
+    clean_text = re.sub(r'<.*?>', '', review)
+    return clean_text
+
+def remove_special(review):
+    cleaned = re.sub(r'[^a-zA-Z0-9\s]', '', review)
+    return cleaned
+
+def remove_stopwords(review):
+    stop_words = set(stopwords.words('english'))
+    words = review.split()
+    filtered_words = [word for word in words if word not in stop_words]
+    return " ".join(filtered_words)
+
+def apply_stemming(review):
+    stemmer = PorterStemmer()
+    words = word_tokenize(review)
+    stemmed_words = [stemmer.stem(word) for word in words]
+    return " ".join(stemmed_words)
+
+def preprocess_text(text):
+    text = remove_html_tags(text)
+    text = text.lower()
+    text = remove_special(text)
+    text = remove_stopwords(text)
+    text = apply_stemming(text)
+    return text
+
+# Load or train model
+@st.cache_resource
+def load_or_train_model():
+    model_file = 'sentiment_model.pkl'
+    vectorizer_file = 'vectorizer.pkl'
+    encoder_file = 'label_encoder.pkl'
+    
+    if os.path.exists(model_file) and os.path.exists(vectorizer_file) and os.path.exists(encoder_file):
+        # Load saved model
+        with open(model_file, 'rb') as f:
+            model_data = pickle.load(f)
+        with open(vectorizer_file, 'rb') as f:
+            cv = pickle.load(f)
+        with open(encoder_file, 'rb') as f:
+            le = pickle.load(f)
+        return model_data['model'], cv, le, model_data['accuracy']
+    else:
+        # Train new model
+        try:
+            df = pd.read_csv("IMDB Dataset.csv")
+            df = df.sample(1200, random_state=42)
+            
+            # Preprocess data
+            df['review'] = df['review'].apply(remove_html_tags)
+            df['review'] = df['review'].str.lower()
+            df['review'] = df['review'].apply(remove_special)
+            df['review'] = df['review'].apply(remove_stopwords)
+            df['review'] = df['review'].apply(apply_stemming)
+            
+            # Vectorize
+            cv = CountVectorizer()
+            X = cv.fit_transform(df['review'])
+            
+            # Encode labels
+            le = LabelEncoder()
+            y = le.fit_transform(df['sentiment'])
+            
+            # Train model
+            from sklearn.model_selection import train_test_split
+            X_train, X_test, y_train, y_test = train_test_split(
+                X, y, test_size=0.2, random_state=42
+            )
+            
+            model = MultinomialNB()
+            model.fit(X_train, y_train)
+            y_pred = model.predict(X_test)
+            accuracy = accuracy_score(y_test, y_pred)
+            
+            # Save model
+            with open(model_file, 'wb') as f:
+                pickle.dump({'model': model, 'accuracy': accuracy}, f)
+            with open(vectorizer_file, 'wb') as f:
+                pickle.dump(cv, f)
+            with open(encoder_file, 'wb') as f:
+                pickle.dump(le, f)
+            
+            return model, cv, le, accuracy
+        except:
+            st.error("Could not load dataset. Please ensure 'IMDB Dataset.csv' is in the same directory.")
+            return None, None, None, None
+
+# Load model
+model, vectorizer, label_encoder, model_accuracy = load_or_train_model()
+
+# Main content
+col1, col2 = st.columns([2, 1])
+
+with col1:
+    st.header("📝 Enter Your Movie Review")
+    
+    # Text input
+    user_input = st.text_area(
+        "Type or paste your movie review here:",
+        height=150,
+        placeholder="E.g., 'This movie was absolutely fantastic! The acting was superb and the plot kept me engaged throughout...'"
+    )
+    
+    # Predict button
+    if st.button("🔍 Analyze Sentiment", type="primary", use_container_width=True):
+        if user_input.strip() == "":
+            st.warning("⚠️ Please enter a movie review first!")
+        elif model is None:
+            st.error("❌ Model not available. Please check the dataset.")
+        else:
+            with st.spinner("Analyzing sentiment..."):
+                # Preprocess input
+                processed_text = preprocess_text(user_input)
+                
+                # Vectorize
+                text_vector = vectorizer.transform([processed_text])
+                
+                # Predict
+                prediction = model.predict(text_vector)[0]
+                sentiment = label_encoder.inverse_transform([prediction])[0]
+                
+                # Get prediction probabilities
+                probabilities = model.predict_proba(text_vector)[0]
+                confidence = max(probabilities) * 100
+                
+                # Display result
+                st.markdown("---")
+                st.subheader("📊 Prediction Result")
+                
+                if sentiment == "positive":
+                    st.markdown(f"""
+                    <div class="prediction-box positive">
+                        <h2>😊 POSITIVE REVIEW</h2>
+                        <p style="font-size: 1.2rem; margin-top: 10px;">
+                            This review expresses a <strong>positive</strong> sentiment about the movie.
+                        </p>
+                        <p style="font-size: 1rem; margin-top: 15px;">
+                            Confidence: <strong>{confidence:.2f}%</strong>
+                        </p>
+                    </div>
+                    """, unsafe_allow_html=True)
+                else:
+                    st.markdown(f"""
+                    <div class="prediction-box negative">
+                        <h2>😞 NEGATIVE REVIEW</h2>
+                        <p style="font-size: 1.2rem; margin-top: 10px;">
+                            This review expresses a <strong>negative</strong> sentiment about the movie.
+                        </p>
+                        <p style="font-size: 1rem; margin-top: 15px;">
+                            Confidence: <strong>{confidence:.2f}%</strong>
+                        </p>
+                    </div>
+                    """, unsafe_allow_html=True)
+                
+                # Show probabilities
+                with st.expander("📈 View Prediction Probabilities"):
+                    prob_df = pd.DataFrame({
+                        'Sentiment': label_encoder.classes_,
+                        'Probability': probabilities
+                    })
+                    prob_df['Probability'] = prob_df['Probability'] * 100
+                    prob_df = prob_df.sort_values('Probability', ascending=False)
+                    st.bar_chart(prob_df.set_index('Sentiment'))
+
+with col2:
+    st.header("ℹ️ About")
+    
+    st.info("""
+    This app analyzes movie reviews and predicts whether the sentiment is **positive** or **negative**.
+    
+    ### How it works:
+    1. Your review is cleaned and preprocessed
+    2. Text is converted to numerical features
+    3. Naive Bayes model makes the prediction
+    4. Results are displayed with confidence score
+    """)
+    
+    if model_accuracy is not None:
+        st.success(f"**Model Accuracy**: {model_accuracy*100:.2f}%")
+    
+    st.markdown("---")
+    
+    st.header("🧪 Try Examples")
+    
+    example_reviews = [
+        "This movie was absolutely fantastic! The acting was superb and the plot kept me engaged throughout.",
+        "Terrible waste of time. Poor acting, boring storyline, and awful special effects.",
+        "A masterpiece of cinema. Brilliant direction and outstanding performances from the entire cast.",
+        "I fell asleep halfway through. The pacing was painfully slow and nothing happened."
+    ]
+    
+    for i, example in enumerate(example_reviews):
+        if st.button(f"Example {i+1}", key=f"ex_{i}"):
+            st.session_state.user_input = example
+
+# Footer
+st.markdown("---")
+st.markdown("""
+<div style="text-align: center; color: #666; font-size: 0.9rem;">
+    <p>Built with ❤️ using Streamlit | Naive Bayes Sentiment Analysis</p>
+    <p>Dataset: IMDB Movie Reviews</p>
+</div>
+""", unsafe_allow_html=True)