In the programming class, teachers need to supervise students as efficiently as they can. Otherwise, they cannot supervise all students. An automatic detection of students who write inappropriate source codes enables teachers to supervise them with higher priority.In this paper, we propose a method to detect backward students. The method calculates the similarity of student codes to model codes in the context of teaching units. The students whose codesare low in the similarity to the model code are regarded as the backward student. The method facilitates teachers to detect students to be supervised more easily.

Full text: IJEIR-1376_final.pdf