diff --git a/backend/config.py b/backend/config.py
index 020f9a9..2531cff 100644
--- a/backend/config.py
+++ b/backend/config.py
@@ -7,6 +7,7 @@ DATA_DIR = os.path.join(BASE_DIR, 'data')
 RAW_DATA_DIR = os.path.join(DATA_DIR, 'raw')
 PROCESSED_DATA_DIR = os.path.join(DATA_DIR, 'processed')
 MODELS_DIR = os.path.join(BASE_DIR, 'models')
+SHAP_CACHE_DIR = os.path.join(MODELS_DIR, 'shap_cache')
 
 RAW_DATA_FILENAME = 'china_enterprise_absence_events.csv'
 RAW_DATA_PATH = os.path.join(RAW_DATA_DIR, RAW_DATA_FILENAME)
diff --git a/backend/core/generate_shap_cache.py b/backend/core/generate_shap_cache.py
new file mode 100644
index 0000000..e74054c
--- /dev/null
+++ b/backend/core/generate_shap_cache.py
@@ -0,0 +1,63 @@
+import json
+import os
+import sys
+
+CURRENT_DIR = os.path.dirname(os.path.abspath(__file__))
+BASE_DIR = os.path.dirname(CURRENT_DIR)
+if BASE_DIR not in sys.path:
+    sys.path.insert(0, BASE_DIR)
+
+import config
+from core.shap_analysis import SHAPAnalyzer
+
+
+DEFAULT_MODELS = ['random_forest']
+
+
+def build_cache(model_type):
+    analyzer = SHAPAnalyzer()
+    global_data = analyzer.global_shap_values(model_type)
+    if global_data.get('error'):
+        raise RuntimeError(global_data['error'])
+
+    top_features = [
+        item['name']
+        for item in global_data.get('top_features', [])[:15]
+    ]
+
+    dependence = {}
+    for feature_name in top_features:
+        data = analyzer.shap_dependence(feature_name, model_type)
+        if not data.get('error'):
+            dependence[feature_name] = data
+
+    interaction = analyzer.shap_interaction(model_type, top_n=10)
+    if interaction.get('error'):
+        raise RuntimeError(interaction['error'])
+
+    return {
+        'model_type': model_type,
+        'global': global_data,
+        'dependence': dependence,
+        'interaction': interaction,
+    }
+
+
+def save_cache(model_type, payload):
+    os.makedirs(config.SHAP_CACHE_DIR, exist_ok=True)
+    cache_path = os.path.join(config.SHAP_CACHE_DIR, f'{model_type}.json')
+    with open(cache_path, 'w', encoding='utf-8') as fp:
+        json.dump(payload, fp, ensure_ascii=False)
+    print(f'Saved SHAP cache: {cache_path}')
+
+
+def main():
+    model_types = sys.argv[1:] or DEFAULT_MODELS
+    for model_type in model_types:
+        print(f'Generating SHAP cache for {model_type}...')
+        payload = build_cache(model_type)
+        save_cache(model_type, payload)
+
+
+if __name__ == '__main__':
+    main()
diff --git a/backend/services/shap_service.py b/backend/services/shap_service.py
index 931e646..f902d94 100644
--- a/backend/services/shap_service.py
+++ b/backend/services/shap_service.py
@@ -1,3 +1,7 @@
+import json
+import os
+
+import config
 from core.shap_analysis import SHAPAnalyzer
 
 
@@ -11,21 +15,60 @@ class SHAPService:
         if self._analyzer is None:
             self._analyzer = SHAPAnalyzer()
 
+    def _get_cache_path(self, model_type):
+        return os.path.join(config.SHAP_CACHE_DIR, f'{model_type}.json')
+
+    def _load_cache(self, model_type):
+        cache_path = self._get_cache_path(model_type)
+        if not os.path.exists(cache_path):
+            return None
+        try:
+            with open(cache_path, 'r', encoding='utf-8') as fp:
+                return json.load(fp)
+        except Exception:
+            return None
+
     def get_global_importance(self, model_type='random_forest'):
-        self._ensure_analyzer()
-        return self._analyzer.global_shap_values(model_type)
+        cache = self._load_cache(model_type)
+        if not cache:
+            return {
+                'error': f'SHAP cache not found for {model_type}. '
+                         f'Run backend/core/generate_shap_cache.py first.'
+            }
+        return cache.get('global', {'error': f'Invalid SHAP cache for {model_type}'})
 
     def get_local_explanation(self, data, model_type='random_forest'):
         self._ensure_analyzer()
         return self._analyzer.local_shap_values(data, model_type)
 
     def get_interactions(self, model_type='random_forest', top_n=10):
-        self._ensure_analyzer()
-        return self._analyzer.shap_interaction(model_type, top_n)
+        cache = self._load_cache(model_type)
+        if not cache:
+            return {
+                'error': f'SHAP cache not found for {model_type}. '
+                         f'Run backend/core/generate_shap_cache.py first.'
+            }
+        data = cache.get('interaction')
+        if not data:
+            return {'error': f'Interaction cache missing for {model_type}'}
+        if top_n and data.get('top_interactions'):
+            result = dict(data)
+            result['top_interactions'] = data['top_interactions'][:top_n]
+            return result
+        return data
 
     def get_dependence(self, feature_name, model_type='random_forest'):
-        self._ensure_analyzer()
-        return self._analyzer.shap_dependence(feature_name, model_type)
+        cache = self._load_cache(model_type)
+        if not cache:
+            return {
+                'error': f'SHAP cache not found for {model_type}. '
+                         f'Run backend/core/generate_shap_cache.py first.'
+            }
+        dependence_map = cache.get('dependence', {})
+        data = dependence_map.get(feature_name)
+        if data:
+            return data
+        return {'error': f'Dependence cache missing for feature {feature_name}'}
 
 
 shap_service = SHAPService()