AND

Hint annotation is one of the most time consuming components of developing intelligent tutoring systems. One approach is to use educational data mining and machine learning techniques to automate the creation of hints from student problem-solving data. This paper describes a new technique to represent, classify, and use programs written by novices as a base for automatic hint generation for programming tutors. Our preliminary evaluation shows that this approach can effectively cluster programs and therefore has potential to be a source for automatically generated hints for novice programmers