Update app.py

app.py

@@ -87,27 +87,71 @@ def get_model():
 @app.post("/classify/")
 def classify_items(request: ItemRequest):
     try:
-        raw_items = request.items_text
+        items_list = request.items
         
-        if not raw_items.strip():
-             return {"success": True, "category_id": None, "category_name": "Tidak Terdeteksi"}
+        # Jika array kosong
+        if not items_list:
+             return {
+                 "success": True, 
+                 "class_id": None, 
+                 "category_name": "Tidak Terdeteksi",
+                 "confidence": 0.0
+             }
              
         m, e = get_model()
         
-        # Inference
-        input_data = tf.constant([raw_items], dtype=tf.string)
-        hasil_prediksi = m.predict(input_data, verbose=0)
+        # Dictionary untuk melacak hasil
+        category_counts = {}
+        max_confidences = {}
         
-        category_id = int(np.argmax(hasil_prediksi[0]))
-        category_name = e.inverse_transform([category_id])[0]
+        # Prediksi satu per satu
+        for item in items_list:
+            if not item.strip():
+                continue
+                
+            input_data = tf.constant([item], dtype=tf.string)
+            hasil_prediksi = m.predict(input_data, verbose=0)
+            
+            cat_id = int(np.argmax(hasil_prediksi[0]))
+            conf = float(np.max(hasil_prediksi[0]))
+            
+            # Hitung frekuensi kategori
+            if cat_id in category_counts:
+                category_counts[cat_id] += 1
+            else:
+                category_counts[cat_id] = 1
+                
+            # Simpan confidence tertinggi untuk kategori
+            if cat_id not in max_confidences or conf > max_confidences[cat_id]:
+                max_confidences[cat_id] = conf
+                
+        # Jika semua item string kosong
+        if not category_counts:
+             return {
+                 "success": True, 
+                 "class_id": None, 
+                 "category_name": "Tidak Terdeteksi",
+                 "confidence": 0.0
+             }
+
+        # Candidate Kategori terbanyak
+        best_cat_id = max(category_counts, key=category_counts.get)
+        best_cat_name = e.inverse_transform([best_cat_id])[0]
+        
+        # Ambil confidence tertinggi dari hasil candidate
+        best_confidence = max_confidences[best_cat_id]
         
         return {
             "success": True, 
-            "class_id": category_id,
-            "category_name": category_name
+            "class_id": best_cat_id,
+            "category_name": best_cat_name,
+            "confidence": round(best_confidence * 100, 2)
         }
         
     except Exception as err:
         import traceback
         print(f"Error AI Klasifikasi: {traceback.format_exc()}")
-        return {"success": False, "error": str(err), "category_id": None, "category_name": None}
+        return {
+            "success": False, "error": str(err), "class_id": None, "category_name": None,
+            "confidence": 0.0
+        }