Mobile Internet is a fast growing, dynamic field, and has wide-ranging application prospects. Electronic commerce is an important application of mobile Internet and it is increasingly changing people’s way of life in the information era. For the present, electronic commerce business is confronted with such challenges as homogeneous profit pattern, customer churn, tenuous loyalty, single channel and so on. Compared to traditional electronic commerce, mobile electronic commerce has incomparable advantages in location, urgency and anytime, anywhere access. On the foundation of association rules in data mining, the architecture of cell phone photograph data mining system and recommendation system model is totally researched. It used the association rule mining algorithm to bulid a cell phone photograph data mining system model. It is very effective for cell phone photograph data mining problem in mobile electronice commerce.