Update app.py

app.py

@@ -1,56 +1,49 @@
 import pickle
 import streamlit as st
-import os
 import numpy as np
 
-# 💡 Define the custom tokenizer exactly as used during training
-def custom_tokenizer(text):
-    # Modify this function to match your original tokenizer logic
-    return text.lower().split()
+# Load saved model, vectorizer, and binarizer
+with open("vectorizer.pkl", "rb") as f:
+    vectorizer = pickle.load(f)
 
+with open("model.pkl", "rb") as f:
+    model = pickle.load(f)
 
+with open("binarizer.pkl", "rb") as f:
+    mlb = pickle.load(f)
 
-# 🔃 Load model files
-try:
-    with open("vectorizer (3).pkl", "rb") as f:
-        vectorizer = pickle.load(f)
-
-    with open("model (6).pkl", "rb") as f:
-        model = pickle.load(f)
-
-    with open("binarizer (3).pkl", "rb") as f:
-        mlb = pickle.load(f)
-
-except Exception as e:
-    st.error(f"❌ Error loading model files: {str(e)}")
-    st.stop()
-
-# 🧠 Prediction function
-def predict_tags(title, description):
-    try:
-        if not title.strip() or not description.strip():
-            return "⚠️ Please enter both title and description."
-
-        input_text = title + " " + description
-        input_vector = vectorizer.transform([input_text])
-        prediction = model.predict(input_vector)
-        predicted_tags = mlb.inverse_transform(prediction)
-        st.write(predicted_tags)
-        if predicted_tags and predicted_tags[0]:
-            return "✅ Predicted Tags: " + ", ".join(predicted_tags[0])
-        else:
-            return "ℹ️ No tags predicted. Try refining your question."
-
-    except Exception as e:
-        return f"❌ Error during prediction: {str(e)}"
-
-# 🚀 Streamlit UI
 st.title("🔖 Stack Overflow Tags Predictor")
-st.markdown("Enter a question title and description to predict relevant tags.")
+st.markdown("Enter a question title and description. Tags will be predicted automatically based on model confidence.")
 
 title = st.text_input("📌 Enter Question Title")
 description = st.text_area("📝 Enter Question Description", height=150)
 
+# Threshold for prediction
+threshold = 0.4  # You can tweak this value to make it more or less strict
+
+def predict_tags_auto(title, description, threshold=0.4):
+    input_text = title + " " + description
+    input_vector = vectorizer.transform([input_text])
+    
+    # Get probabilities for each tag
+    probas = model.predict_proba(input_vector)
+    
+    # Convert list of (1, n_classes) probs to array
+    probas = np.array([p[0] for p in probas])
+    
+    # Apply threshold
+    predicted_binary = (probas >= threshold).astype(int)
+    
+    # Convert binary vector to tags
+    tags = mlb.inverse_transform(predicted_binary)
+    return tags[0] if tags else []
+
 if st.button("Predict Tags"):
-    result = predict_tags(title, description)
-    st.markdown(result)
+    if not title.strip() or not description.strip():
+        st.warning("⚠️ Please enter both title and description.")
+    else:
+        tags = predict_tags_auto(title, description)
+        if tags:
+            st.success("✅ Predicted Tags: " + ", ".join(tags))
+        else:
+            st.info("ℹ️ No tags predicted. Try refining your question.")