Large amount of users’ trajectories, is an emerging source of inexpensive data that can be used to provide an opportunity to present route recommendation service to the unfamiliar users within the area. In this study, with the aim of finding the optimal route, we first extract both local users and local road segments data sets by ranking them via HITS algorithm. In this model, a hub is a user who many time has crossed many road segments of a region, and an authority is a road segment that has been crossed by many users. Therefore, users’ travel experiences (hub scores) and the interests of road segments (authority scores) have a mutual reinforcement relation. We also propose a novel approach in which the basic unit of routing is separate road segment instead of GPS trajectory segment. Moreover, to provide the approximate routing, we create a local graph. The center of the local road segments are considered as nodes and are based on local streets sequence arrange the pieces obtained from the trajectory of the user as edges of local graph. According to this graph, two steps of routing are used to obtain the optimal path. Then using Dijkstra's algorithm on the main road network and obtained an approximate route, shortest route between two local road segments based on this graph is used to obtain the optimal route. To implement and test, used data, from the trajectories of moving users in Tehran, has been gathered for 3 months on daily basis. To evaluate performance of the two-step routing, we experimentally compared the travel time in proposed route to Dijkstra’s shortest path for different lengths and users with different levels of regional knowledge. The travel time in the proposed method was decreased 60 percent  compare to shortest route.