--- a
+++ b/src/cnn/cnn_utils.py
@@ -0,0 +1,27 @@
+import numpy as np
+
+
+def character_index(sentence, vocabulary, sequence_length = 500):
+  index_list = []
+  for i in range(len(sentence)):
+    if i > sequence_length-1:
+      break
+    else:
+      if sentence[i] in vocabulary:
+        index_list.append(vocabulary.index(sentence[i]) + 1)
+      else :
+        index_list.append(len(vocabulary)+1)
+  if len(index_list) == sequence_length:
+    return index_list
+  else:
+    index_list.extend([0]*(sequence_length-len(index_list)))
+    return index_list
+
+def character_embedding(index_list, vocabulary):
+  embedding_weights = []
+  for index,i in enumerate(index_list):
+    one_hot = np.zeros(len(vocabulary)+1)
+    if i != 0:
+      one_hot[i-1] = 1
+    embedding_weights.append(one_hot)
+  return np.array(embedding_weights,dtype = 'float32').T
\ No newline at end of file