We proposed a collaborative location recommendation system with dynamic time periods. Dynamically assign an expected factor for different POIs to promote recommendation precision. The method filters the POIs with time period threshold. To make this simulation more realistic, we assign different time period threshold to different areas.