app.py

import numpy as np
import tensorflow as tf
from tensorflow.keras.datasets import imdb
from tensorflow.keras.preprocessing import sequence
from tensorflow.keras.models import load_model
import streamlit as st

# Load the IMDB dataset word index
word_index = imdb.get_word_index()
reverse_word_index = {value: key for key, value in word_index.items()}

# Load the pre-trained model with ReLU activation
model = load_model('simple_rnn_imdb.h5')

# Helper function to decode reviews
def decode_review(encoded_review):
    return ' '.join([reverse_word_index.get(i - 3, '?') for i in encoded_review])

# Function to preprocess user input
def preprocess_text(text, max_word_index=10000, maxlen=500):
    words = text.lower().split()
    encoded_review = [min(word_index.get(word, 2) + 3, max_word_index - 1) for word in words]
    padded_review = sequence.pad_sequences([encoded_review], maxlen=maxlen)
    return padded_review

# Streamlit app
st.title('IMDB Movie Review Sentiment Analysis')
st.write('Enter a movie review to classify it as positive or negative.')

# User input
user_input = st.text_area('Movie Review')

if st.button('Classify'):
    preprocessed_input = preprocess_text(user_input)

    # Make prediction
    prediction = model.predict(preprocessed_input)
    sentiment = 'Positive' if prediction[0][0] > 0.5 else 'Negative'

    # Display the result
    st.write(f'Sentiment: {sentiment}')
    st.write(f'Prediction Score: {prediction[0][0]}')
else:
    st.write('Please enter a movie review.')