Nowadays, a rapid development in the communication technology and increasing the usability of powerful portable devices, mobile users can use their mobile devices to access the information. One of the active areas is the mining and prediction of users' mobile commerce behaviors such as their movements and purchase transactions. The important issue is to mine the rare frequent items from database to satisfy the user needs. In this paper, we propose a technique that can efficiently satisfy the user needs. It predicts the frequent item based on the user selection. Systolic tree implementation is used to predict the frequently moved item in the database. The main aim is to recommend the stores and items previously to unknown user. We evaluate our system in real world and deliver good performance in terms efficiency and scalability.