Research On The Prediction Of Railway Travel Group Relationship Based On Frequent Subgraph Mining

In recent years,with the rapid development of the railway passenger transport industry and information technology,the comfort of railway travel and the speed of online ticket purchase have been greatly improved,and railway travel has gradually become the first choice for people to travel.With the accumulation of the number of railway trips year by year,the phenomenon of traveling together becomes more and more obvious.People often travel with others for the same purpose,such as traveling with family,traveling for work with colleagues,or going on vacation with friends.Different types of travel groups have different travel preferences,and their travel behaviors are closely related to them.Accurate prediction of the relationship between railway travel groups can not only provide support for personalized services and product recommendations for transportation,tourism,hotel and other industries,but also provide support for market decision-making in the railway passenger transport industry.Based on this,this paper proposes a method to predict the co-travel relationship of railway travel groups by using the popular frequent subgraph mining algorithm in the graph domain according to the historical co-travel network of railway passengers.Divide group relationships into three categories: family,colleagues,and friends.Firstly,the selection of group characteristic attributes is studied,and the virtual sample generation method is applied to sample amplification for small batches of data based on group attributes.Then,the railway passenger co-travel network is constructed based on the historical ticket purchase records in the railway passenger transportation system.Secondly,a frequent subgraph mining model based on railway travel groups is constructed to mine all frequent travel subgraphs.Finally,based on the frequent subgraph sets,a graph classification model is constructed using the partial order relation topology to classify the group relationship of railway travel.This paper conducts experiments on the accuracy of the proposed prediction method in a real railway ticketing data set.The experimental results show that the accuracy of the proposed method in predicting the relationship between railway travel group is up to0.67.This proves that the prediction model proposed in this paper can effectively and accurately predict the potential co-travel relationship in the railway travel group.