Mathematical exercises consist of several sub-exercises. When studying mathematics using an example exercise (example), some learners memorize all theorems used in the example, while others consider unknown derivation steps logically and memorize only new theorems. The latter learning method is effective, since the amount of memorization is small and learners can consider the reasons for applying newly appearing theorems. The aim of this research is to construct a learning environment that supports learners in generating sub-exercises from an example in order to foster logical thinking about mathematics. By finding known theorems from the example, learners are able to notice relations between exercises. In our environment, suggestions are provided if learners are not able to derive the sub-exercises. In addition, examples are given in the appropriate order to allow learners to easily notice the known theorems.