Update app.py

app.py

@@ -105,6 +105,45 @@ def gru_predict_movie_sentiment(review):
         return "Negative"
 
 
+with open('perceptron_movie_model.pkl', 'rb') as file:
+    perceptron_movie_model = pickle.load(file)
+
+def predict_movie_sentiment_perceptron(review):
+    max_review_length = 500
+    top_words = 5000
+    word_index = imdb.get_word_index()
+    review = review.lower().split()
+    review = [word_index[word] if (word in word_index and word_index[word] < top_words) else 0 for word in review]
+    review_bin = np.where(np.array(review) > 0, 1, 0)
+    # Padding or truncating the review to match the perceptron's input size
+    review_bin_padded = np.pad(review_bin, (0, max_review_length - len(review_bin)), 'constant')
+    prediction = perceptron_movie_model.predict([review_bin_padded])
+    if prediction[0] == 1:
+        return "Positive"
+    else:
+        return "Negative"
+
+        
+# Load the saved instance of the Perceptron class
+with open('backprop_movie_model.pkl', 'rb') as file:
+    backprop_movie_model = pickle.load(file)
+
+def predict_movie_sentiment_backprop(review):
+    max_review_length = 500
+    top_words = 5000
+    word_index = imdb.get_word_index()
+    review = review.lower().split()
+    review = [word_index[word] if (word in word_index and word_index[word] < top_words) else 0 for word in review]
+    review_bin = np.where(np.array(review) > 0, 1, 0)
+    # Padding or truncating the review to match the perceptron's input size
+    review_bin_padded = np.pad(review_bin, (0, max_review_length - len(review_bin)), 'constant')
+    prediction = backprop_movie_model.predict([review_bin_padded])
+    if prediction[0] == 1:
+        return "Positive"
+    else:
+        return "Negative"
+
+
 
 # Main function for Streamlit app
 def main():    
@@ -182,7 +221,27 @@ def main():
                 else:
                     st.write("Please enter a movie review for sentiment analysis")
 
-        
+        elif model == "Perceptron":
+            st.subheader("Movie Sentiment Analysis")
+            user_review = st.text_area("Enter a movie review: ")
+            
+            if st.button("Analyze Sentiment"):
+                if user_review:
+                    sentiment_result = predict_movie_sentiment_perceptron(user_review)
+                    st.write(f"The sentiment of the review is: {sentiment_result}")
+                else:
+                    st.write("Please enter a movie review for sentiment analysis")
+                    
+        elif model == "Backpropagation":
+            st.subheader("Movie Sentiment Analysis")
+            user_review = st.text_area("Enter a movie review: ")
+            
+            if st.button("Analyze Sentiment"):
+                if user_review:
+                    sentiment_result = predict_movie_sentiment_backprop(user_review)
+                    st.write(f"The sentiment of the review is: {sentiment_result}")
+                else:
+                    st.write("Please enter a movie review for sentiment analysis")
 
                 
 if __name__ == "__main__":