Add text classification model to detect spam messages

src/apps/pages/models/SPAM detection model/spam_detector.py

@@ -0,0 +1,86 @@
+import os
+import streamlit as st
+import tensorflow as tf
+import pickle
+from tensorflow.keras.preprocessing.sequence import pad_sequences
+
+# ---------------------------
+# Download model/tokenizer from Kaggle Kernel
+# ---------------------------
+@st.cache_resource
+def download_from_kaggle(output_dir="kaggle_models"):
+    try:
+        os.makedirs(output_dir, exist_ok=True)
+        kernel_ref = "smritipandey02/spam-detection"  # Hardcoded username + notebook
+
+        # Download kernel outputs
+        os.system(f"kaggle kernels output {kernel_ref} -p {output_dir}")
+
+        st.success(f"Downloaded model/tokenizer from Kaggle Kernel: {kernel_ref}")
+        return output_dir
+    except Exception as e:
+        st.error(f"Could not download from Kaggle: {e}")
+        st.stop()
+
+# ---------------------------
+# Load model & tokenizer
+# ---------------------------
+@st.cache_resource
+def load_model_and_tokenizer():
+    try:
+        model_dir = download_from_kaggle()
+
+        model = tf.keras.models.load_model(os.path.join(model_dir, "spam_classifier.h5"))
+        with open(os.path.join(model_dir, "tokenizer.pkl"), "rb") as f:
+            tokenizer = pickle.load(f)
+
+        return model, tokenizer
+    except Exception as e:
+        st.error(f"Could not load model/tokenizer: {e}")
+        st.stop()
+
+# ---------------------------
+# Prediction function
+# ---------------------------
+def predict_message(message, model, tokenizer, max_len=100):
+    seq = tokenizer.texts_to_sequences([message])
+    padded = pad_sequences(seq, maxlen=max_len, padding="post", truncating="post")
+    prob = float(model.predict(padded)[0][0])
+    return prob
+
+# ---------------------------
+# Streamlit UI
+# ---------------------------
+def spam_app():
+    st.title("Spam Message Detector")
+
+    # Load Model button
+    if st.button("Load Model", key="load_btn"):
+        with st.spinner("Loading model & tokenizer..."):
+            model, tokenizer = load_model_and_tokenizer()
+            st.session_state["model"] = model
+            st.session_state["tokenizer"] = tokenizer
+        st.success("Model & Tokenizer Loaded Successfully!")
+
+    # Input box
+    message = st.text_area("Enter a message:", key="msg_input")
+
+    # Predict button
+    if st.button("Predict", key="predict_btn"):
+        if not message.strip():
+            st.warning("Please enter a message before predicting.")
+        elif "model" not in st.session_state or "tokenizer" not in st.session_state:
+            st.warning("Please load the model first.")
+        else:
+            prob = predict_message(message, 
+                                   st.session_state["model"], 
+                                   st.session_state["tokenizer"])
+            if prob > 0.5:
+                st.error(f"Spam detected! (Confidence: {prob:.2f})")
+            else:
+                st.success(f"Ham (Not Spam) (Confidence: {1-prob:.2f})")
+
+# ---------------------------
+# Run app
+# ---------------------------
+spam_app()