diff --git a/memory.py b/memory.py
index 3a283b8..88216fc 100644
--- a/memory.py
+++ b/memory.py
@@ -1,5 +1,6 @@
 import csv
 import json
+from math import sqrt
 
 WORD_THRESHOLD = 4
 
@@ -52,7 +53,7 @@ We define a similarity measure on sets which counts the number of elements
 they have in common
 """
 def scalar(a, b):
-    return len(a.intersection(b))
+    return len(a.intersection(b))/sqrt(len(a)*len(b))
 
 def find_best_quote(db, user_input):
     indexed_input = index(user_input)