Dynamic SPT Increment Algorithm for Aircraft Booking Based on Hadoop

Abstract

In the civil aviation industry, the number of flights and the number of airports have been increasing. Airlines recommendation service which to provide customers with more efficient and shorter routes has become one of the effective measures for improving the service quality of aviation. Actually, Airports and air-lines can be formed a weighted directed graph (for example, we use plane fares to represent the weights on the edges of the graph, and represent the vertices in the graph by the airports), and Airlines recommendation is to help user to find out the shortest path from the source vertex to the target vertices of the weighted directed graph. The shortest path search can be done by using the Dijkstra incremental optimization algorithm[1]. However, as the incremental part become larger and larger, the time spent on shortest path computation will be growing rapidly. In order to shorten the computing time, this paper proposes a dynamic SPT increment algorithm based on Hadoop platform. By the Hadoop efficient distributed platform[2], using MapReduce to take a dynamic search for the incremental part of the SPT. Finally, regenerates the new shortest path tree to improve the algorithm efficiency and performance.