Research On The Discovery Of Aviation Passenger Value Based On Social Network

With the deepening of the informalization in civil aviation, the system has accumulated a large number of passenger information data, it is an urgent task how to effectively use these data and service in the civil aviation industry. The traditional value analysis of civil aviation passenger focuses on the individual traveler, ignoring the relationship between passengers,however, in real life, the interaction between passengers is existed. The research of social network is to consider the value of each node but also quantitative analysis the relationships between nodes. Combined with the latest research achievements of social network mining and analysis field, through the analysis of the internal relationship between passengers, building social networks of civil aviation passengers and digging out the valuable passengers, It could provide decision support for enterprise marketing, help the civil aviation enterprise improve the quality of service, as the same time, the passengers get more targeted and quality aviation services.Firstly, the dissertation proposes a method based on the relationship between the PNR data to build the social network of civil aviation passengers.And experiments with the use of passenger travel records in a lot of routes of an airline, extracting the maximum connected sub-graph in the network, provide favorable experimental basis for further research on the value of network nodes(the critical nodes).Secondly, in order to more accurately find out valuable nodes(important nodes) in network, on the basis of studying and comparing a variety of models, the dissertation proposes a model that discovery the value of social network of aviation passenger based on multi-level and gray relational analysis. Considering the average clustering coefficient cannot measure the clustering scale, so propose the amended clustering coefficient. A Multi-level gray relational analysis model is established to identify the correlation degree between nodes and the optimal nodes,realizing to rank the network nodes. The model is applied in the aviation passenger of social network, get the ideal results, by contrasting with the actual situation, further verify the reliability of the model.Finally, the dissertation presents a value model of aviation passenger of the social networkbased on the important node discovery algorithms. Study the four representative algorithms in the three areas of system science, networking and Internet searching, and comparing the calculated results by F-measure of method, find the important node, And analysis the reliability of nodes by degrees, double degrees, the right of points, double the right of points,finding out the important nodes of each connected sub-graph.In summary, the dissertation mainly build the social network of civil aviation passengers based on PNR data, proposes the model based on multi-level gray relational analysis, the value model of aviation passenger of the social network based on the found algorithm of important nodes. The simulation of the model shows that it can find valuable passengers,reach the purpose of helping airlines make decision, Through the specialized marketing for high-value travelers, get maximum benefits with minimum cost, thereby saving costs for airlines.