GPS-equipped taxis uses the sensor device to detect the road surface, and taxi drivers are usually experienced with  finding the fastest  route to reach the destination based on their knowledge. This takes much time and rough routes are generated. In this paper, we process the  smart driving directions from the historical dataset collected from  large number of taxis. In our approach, we propose a time-dependent landmark graph, where the landmark is denoted as node is a road segment frequently traversed by taxis, to model the intelligence of taxi drivers. We have used a Variance-Entropy-Based Clustering approach to estimate the distribution of travel time between source and destination in different time slots. We also calculate the distance taken to travel the location. Based on this,  we design a two-stage routing algorithm to compute the practically fastest route. We build our project based on a real-world trajectory dataset generated by over 33,000 taxis in a period of 3 months and evaluate the system by conducting both synthetic experiments and in-the-field evaluations. As a result, 60-70% of the routes suggested by our method are faster than the competing methods, and 20% of the routes share the same results. On average, 50% of our routes are at least 20% faster than the competing approaches.