PranavReddy18/Email_Spam_Classification

1import streamlit as st
2import tensorflow as tf
3from tensorflow.keras.preprocessing.text import Tokenizer
4from tensorflow.keras.preprocessing.sequence import pad_sequences
5
6# Load your saved model and tokenizer
7def load_model_and_tokenizer():
8    # Assuming the model is saved as 'spam_ham_model.h5' and tokenizer saved as 'tokenizer.pickle'
9    model = tf.keras.models.load_model('model.h5')
10
11    # You need to have a way to load the tokenizer that you used
12    import pickle
13    with open('tokenizer.pkl', 'rb') as handle:
14        tokenizer = pickle.load(handle)
15    
16    return model, tokenizer
17
18# Preprocessing function for the user input
19def preprocess_input(texts, tokenizer, maxlen=50):
20    sequences = tokenizer.texts_to_sequences(texts)
21    return pad_sequences(sequences, maxlen=maxlen, padding='post')
22
23# Prediction function
24def predict_text(model, tokenizer, sample_texts, maxlen=50):
25    X_predict = preprocess_input(sample_texts, tokenizer, maxlen)
26    predictions = model.predict(X_predict)
27    
28    results = []
29    for text, pred in zip(sample_texts, predictions):
30        label = "spam" if pred[0] > 0.5 else "ham"
31        results.append({
32            "Text": text,
33            "Predicted Label": label,
34            "Prediction Confidence": f"{pred[0]:.4f}"
35        })
36    return results
37
38# Streamlit App Interface
39def main():
40    st.title('Spam vs Ham Text Classifier')
41    st.markdown("""
42    This is a simple Streamlit app that predicts whether a given text is **Spam** or **Ham** using a pre-trained model.
43    """)
44
45    # Load model and tokenizer
46    model, tokenizer = load_model_and_tokenizer()
47
48    # Text input
49    text_input = st.text_area("Enter the text you want to classify:")
50
51    # Button to predict
52    if st.button("Predict"):
53        if text_input:
54            # Get the prediction
55            prediction_results = predict_text(model, tokenizer, [text_input])
56            
57            # Display the result
58            for result in prediction_results:
59                st.write(f"**Text**: {result['Text']}")
60                st.write(f"**Predicted Label**: {result['Predicted Label']}")
61                st.write(f"**Prediction Confidence**: {result['Prediction Confidence']}")
62        else:
63            st.error("Please enter some text to classify.")
64
65# Run the app
66if __name__ == "__main__":
67    main()
68