Conclusions and future Research
A given path in a graph can be split in Basic Path Components (BPCs) that are either least-cost paths or non-least-cost edges. We developed an algorithm that computes the size of the minimum path splitting in an efficient way. A graph-theoretical correctness proof for the algorithm is given. This analysis sheds a light on the characteristics of splitVertexSuites (subpaths consisting of potential splitVertices) and their relationship to specific minimal shortcuts. It results in a set of splitVertexSuites from which minimum path splittings can be generated and an upper bound for the number of possible minimum path splittings. This algorithm is used to verify the hypothesis that for utilitarian trips, people tend to compose their route from a small number of least cost paths. Both car and person GPS traces have been recorded. Individual trips have been identified in