Combined-cost and bi-objective approaches to multipurpose location-based services using genetic algorithms

This paper discusses a multipurpose location-based service providing users with integrated information on transportation and points of interest (POIs) to support smarter decisions regarding route choice. The problem is finding a route and location of POI between an origin and a destination with minimum total cost including the purchasing and travel costs. To respond to a user's query for such a service, a solution algorithm should be both efficient and accurate enough to maintain serviceability and commerciality. The combined-cost and bi-objective approaches are proposed, where the former is to use the time value to combine the travel and purchasing costs, and the latter divides two objectives and finds the multiple reasonable solutions for both objectives. The bi-objective approach may take additional time and resources for the multiple solutions but can provide answers for more complex service requests even when the time value is not known. Different versions of hybrid genetic algorithms are suggested for both approaches and tested in a large-size Chicago, Illinois, transportation network.