Comparison of Three Search Algorithms for Mobile Trip Planner for Eskisehir City

Abstract

The comparison of three searching algorithms; A*, Ant Colony Optimization and Genetic Algorithms to solve the Traveler Salesman Problem for a mobile trip planning application for Eskisehir City, Turkey, is presented in this paper. The algorithms work on more than 30 point-of-interests and 150 sub point-of-interests. The algorithms are compared with respect to their running times for scenarios with different number of point-of-interests. Experimental results show that the A* algorithm is 400-600% faster than the other algorithms. The mobile application calculates the best route trip planned according to the traveler's preferences on categorized points-of-interests. The mobile application also recommends alternative route plans during the trip when the traveler is ahead or behind the schedule.