The value of offline data is continuously realized due to the rapid development of GPS, GSM and Wireless Sensor Technology. The Mobile Internet has become more common and with the advancement of technology, people have become more dependent on the mobile intelligent terminal. The Advertising model is influenced subtly by changes to the audiences’ lives and online patterns. Location Based Social Networking Services promote the precision advertising model into being, which is based on mobile internet. This paper used the methods of trajectory mining and POI clustering to create activity mapping table based on the users, location track data, then established user behavior feature model and built an advertising platform to implement precision advertising.