Mobile Commerce, also known as M-Commerce or mCommerce, is the ability to conduct commerce using a mobile device. Research is done by Mining and Prediction of Mobile Users’ Commerce Behaviors such as their purchase transactions. The problem of PMCP-Mine algorithm has been overcome by the efficient framework based graph diffusion method. The main objective is to construct the graph based diffusion method. Graph is constructed for the items purchased by the Mobile users and then finding the frequently purchased item. By using ranking method, we are ranking the items based on the transactions. Then, by analyzing the mobile users behavior and recommending the ranked items. This framework produces more efficient and accurate item recommendation than the MCE framework.