diff --git a/README.md b/README.md
index ab70568..f67e012 100644
--- a/README.md
+++ b/README.md
@@ -9,8 +9,9 @@ For each potential rule, the rule validator will take as inputs :
   - a file containing the "bad code" (code smell)
   - a file containing the test scenarios to be executed.
 
-  For each scenario, the rule validator will instrumente it with monitoring tools (actually only one - vjoule) and will generate some results file.
+  For each scenario, the rule validator will instrumente it with monitoring tools (actually only one - vjoule) and will generate some results file (csv files).
 
   You can find some generated sample files in the folder **/samples**. 
 
-  Based on those results file, a graphic summary is generated for each rule.
+  Based on those results files, a graphic summary is generated for each rule using 'graph_generator.py'.
+  'graph_generator.py' takes as input the previously generated csv results file, and based on that file, an html page containing the comparison graph is generated. Note that this script takes the name of the compared function as an input argument. the csv file name AND the first cell of its content must be that function name.