The increasing availabilities of GPS-enabled devices have given rise to the location-based social networking services (LBSN), in which users can record their travel experiences with GPS trajectories and share these trajectories among each other on Web communities. Usually, GPS-enabled devices record far denser points than necessary in the scenarios of GPS-trajectory-sharing. Meanwhile, these redundant points will decrease the performance of LBSN systems and even cause the Web browser crashed. Existing line simplification algorithms only focus on maintaining the shape information of a GPS trajectory while ignoring the corresponding semantic meanings a trajectory implies. In the LBSN, people want to obtain reference knowledge from other users’ travel routes and try to follow a specific travel route that interests them. Therefore, the places where a user stayed, took photos, or changed moving direction greatly, etc, would be more significant than other points in presenting semantic meanings of a trajectory. In this paper, we propose a trajectory simplification algorithm (TS), which considers both the shape skeleton and the semantic meanings of a GPS trajectory. The heading change degree of a GPS point and the distance between this point and its adjacent neighbors are used to weight the importance of the point. We evaluated our approach using a new metric called normalized perpendicular distance. As a result, our method outperforms the DP (Douglas-Peuker) algorithm, which is regarded as the best one for line simplification so far.