diff --git a/src/model/q_a.py b/src/model/q_a.py
index 3a97829c2561b65ab4061640b5f61506e1619f2d..e9b8968641654e7f5e8fd87965e82007f8761ac0 100644
--- a/src/model/q_a.py
+++ b/src/model/q_a.py
@@ -10,16 +10,20 @@ context.set_context(mode=context.GRAPH_MODE, device_target="CPU")
 
 
 def compute_similarity(input_encode):
-    max_similarity = 0
     with open("../data/resource_sentence_encode.json") as f:
         resource_sentences_encode = json.load(f)
+        question_similarity = {}
         for k, v in resource_sentences_encode.items():
             similarity = cosine_similarity(
                 [input_encode[1][0].asnumpy()], [np.asarray(v)])
-            if similarity > max_similarity:
-                max_similarity = similarity
-                match_key = k
-    return match_key
+            question_similarity[k] = similarity
+        sorted_similarity = sorted(question_similarity.items(), key=lambda x: x[1], reverse=True)
+        if not sorted_similarity or len(sorted_similarity) == 0:
+            return None
+        elif sorted_similarity[0][1] >= 0.7:
+            return [sorted_similarity[0][0]]
+        elif len(sorted_similarity) > 1 and sorted_similarity[0][1] < 0.7:
+            return [sorted_similarity[0][0], sorted_similarity[1][0]]
 
 
 def encode_sentence(input_sentence):
@@ -38,7 +42,9 @@ def load_q_a_data(file_path):
 if __name__ == '__main__':
     input_sentence = sys.argv[1]
     input_encode = encode_sentence(input_sentence)
-    match_key = compute_similarity(input_encode)
+    match_keys = compute_similarity(input_encode)
     q_a_data = load_q_a_data("../data/q_a.json")
-    if match_key in q_a_data.keys():
-        print(q_a_data[match_key])
+    if match_keys:
+        for k in match_keys:
+            if k in q_a_data.keys():
+                print(q_a_data[k])
diff --git a/src/qaRobot/qa_api/qa/q_a.py b/src/qaRobot/qa_api/qa/q_a.py
index 2f55a9ab122daccdbe08bc52f1b6efcf023b0982..9044cd4741ee9cba49d03c0f03ae585357ee859c 100644
--- a/src/qaRobot/qa_api/qa/q_a.py
+++ b/src/qaRobot/qa_api/qa/q_a.py
@@ -12,17 +12,20 @@ module_dir = os.path.dirname(__file__)
 
 
 def get_match_question(input_encode):
-    max_similarity = 0
-    print(os.path.join(os.path.dirname(os.path.dirname(__file__))))
     with open(os.path.join(module_dir, '../data/resource_sentence_encode.json')) as f:
         resource_sentences_encode = json.load(f)
+        question_similarity = {}
         for k, v in resource_sentences_encode.items():
             similarity = cosine_similarity(
                 [input_encode[1][0].asnumpy()], [np.asarray(v)])
-            if similarity > max_similarity:
-                max_similarity = similarity
-                match_key = k
-    return match_key
+            question_similarity[k] = similarity
+        sorted_similarity = sorted(question_similarity.items(), key=lambda x: x[1], reverse=True)
+        if not sorted_similarity or len(sorted_similarity) == 0:
+            return None
+        elif sorted_similarity[0][1] >= 0.7:
+            return [sorted_similarity[0][0]]
+        elif len(sorted_similarity) > 1 and sorted_similarity[0][1] < 0.7:
+            return [sorted_similarity[0][0], sorted_similarity[1][0]]
 
 
 def encode_sentence(input_sentence):
diff --git a/src/qaRobot/qa_api/views.py b/src/qaRobot/qa_api/views.py
index 37b39d4d89064b2787c181c74f48360d221e2cbb..8958aade0af7887614c69505bae9a16bc3ea1b4d 100644
--- a/src/qaRobot/qa_api/views.py
+++ b/src/qaRobot/qa_api/views.py
@@ -13,12 +13,16 @@ module_dir = os.path.dirname(__file__)
 def get_answer(request):
     question = request.query_params.get('q', None)
     input_encode = encode_sentence(question)
-    match_key = get_match_question(input_encode)
+    match_keys = get_match_question(input_encode)
     file_path = os.path.join(module_dir, 'data/q_a.json')
     q_a_data = {}
     if not q_a_data:
         q_a_data = load_q_a_data(file_path)
-    r = {"question": question}
-    if match_key in q_a_data.keys():
-        r["answer"] = q_a_data[match_key]
+    r = []
+    if match_keys:
+        for k in match_keys:
+            q_a = {"question": k}
+            if k in q_a_data.keys():
+                q_a["answer"] = q_a_data[k]
+                r.append(q_a)
     return HttpResponse(json.dumps(r, ensure_ascii=False), content_type='application/json')