I. INTRODUCTION
There are many methods in programming education based on automated testing, in which submitted programs are built and tested automatically [1]. Evaluation in such methods is performed by whether they passed the tests teachers defined in advance. We call such methods as test-based education in this paper. Test-based education has a certain effect on motivating students for programming since the success or failure of their assignment can be determined automatically and immediately. It is also capable of reducing the burden on teachers because they no longer need to check all the assignments [1].