Research On Civil Aviation Passenger Influence Calculation Based On Social Network

There is a “viral marketing” strategy in marketing field,which requires to select a few influential users in the user group with limit resource,and then use social network to spread marketing information to more user in a short period of time.This problem is formalized as the influence maximization problem.In the civil aviation passenger group,there are also some influential passengers who play a crucial role in the marketing and promotion of airlines.This paper mainly utilizes the knowledge of social network to excavate the relationship among passengers in the passenger reservation data and construct the civil aviation passenger social network.Then,mining the most influential passengers in the passenger group based on this network.The specific contents are as follows:1.A method of parallel construction of civil aviation passenger social network based on Passenger Name Record data is proposed by analyzing PNR data.The passengers are token as nodes and the edge between nodes is defined by mining passengers' common flight and booking behavior.The weights of edges are measured by common trip confidence between passengers.The network construction method proposed in this paper is proved to be effective and feasible through experiments on the airline's real booking data set.2.Aiming at the calculation of civil aviation passenger influence,a new civil aviation passenger influence calculation model called CAPI-Rank is proposed,which combines the network topology and passenger behavior.Then a new influence maximization algorithm called BCPIG algorithm is proposed based on the effect of weak ties between communities on information diffusion and CAPI-Rank model.The algorithm is validated experimentally on social networks with different scales to verify the performance of the algorithm and is compared with Greedy algorithm and DegreeDiscount algorithm in algorithmic precision and algorithmic efficiency.The experimental results show that the BCPIG algorithm greatly reduces the time complexity compared with Greedy algorithm,and it is better than DegreeDiscount heuristic algorithm in algorithmic precision.