Update app.py

app.py

@@ -1,18 +1,37 @@
 import streamlit as st
 from transformers import pipeline
 
-# Load emotion classification pipeline
+# Load models
 emotion_classifier = pipeline(
     "text-classification",
     model="bhadresh-savani/distilbert-base-uncased-emotion",
-    top_k=3  # Return top 3 emotions
+    top_k=3
 )
 
-# Define emotion-based priority rules
+intent_classifier = pipeline(
+    "zero-shot-classification",
+    model="facebook/bart-large-mnli"
+)
+
+# Define emotion priority rules
 urgent_emotions = {"anger", "annoyance", "disgust", "frustration", "sadness"}
 moderate_emotions = {"confusion", "concern", "nervousness", "fear"}
 low_emotions = {"neutral", "approval", "excitement", "joy", "curiosity"}
 
+# Define candidate customer intents
+candidate_tasks = [
+    "change data plan",
+    "upgrade phone",
+    "top up balance",
+    "report network issue",
+    "ask for billing help",
+    "request human support",
+    "check account status",
+    "suspend service",
+    "reactivate number",
+    "cancel subscription"
+]
+
 def assess_priority(emotion):
     if emotion in urgent_emotions:
         return "🔴 High", "✅ Immediate human support is recommended."
@@ -21,33 +40,38 @@ def assess_priority(emotion):
     else:
         return "🟢 Low", "❌ No human support needed. Automated response is sufficient."
 
-# Streamlit page settings
-st.set_page_config(page_title="AI Customer Emotion Analyzer", layout="centered")
-st.title("📞 AI Customer Emotion & Priority Analyzer")
+# Streamlit App Interface
+st.set_page_config(page_title="AI Customer Support Analyzer", layout="centered")
+st.title("📞 AI Customer Emotion & Intent Analyzer")
 
-# User input
 user_input = st.text_area("Please enter the customer's message or conversation:", height=150)
 
-# Analyze button
-if st.button("Analyze Emotion"):
+if st.button("Analyze"):
     if user_input.strip() == "":
         st.warning("Please enter a message to analyze.")
     else:
-        with st.spinner("Analyzing emotion..."):
-            # Run emotion analysis
+        with st.spinner("Analyzing..."):
+
+            # --- Emotion Classification ---
             emotion_results = emotion_classifier(user_input)
             top_emotion = emotion_results[0][0]
             emotion_label = top_emotion['label']
             emotion_score = top_emotion['score']
-
-            # Determine priority level
             priority_level, recommendation = assess_priority(emotion_label)
 
-            # Display results
-            st.subheader("📌 Emotion Analysis Results")
+            st.subheader("📌 Emotion Analysis")
             st.write(f"**Primary Emotion**: {emotion_label}")
             st.write(f"**Confidence Score**: {emotion_score:.2f}")
 
             st.subheader("🛎️ Support Priority Recommendation")
             st.write(f"**Priority Level**: {priority_level}")
             st.success(recommendation)
+
+            # --- Intent Detection ---
+            task_result = intent_classifier(user_input, candidate_tasks)
+            top_tasks = task_result['labels'][:3]
+            top_scores = task_result['scores'][:3]
+
+            st.subheader("⚙️ Detected Possible Customer Intents")
+            for label, score in zip(top_tasks, top_scores):
+                st.write(f"🔸 **{label}** (confidence: {score:.2f})")