Abstract: Big data provide a huge amount of information and bring the information overload problem to users.This situa-tion is more serious in tourism field.To solve the problem that tourists need to spend a lot of time and energy to prepare a travel plan,this paper firstly proposed a method to construct a knowledge graph incorporating heterogeneous tourism data to extract the knowledge in tourism field.Secondly,this paper used the knowledge graph and travelogues to gene-rate the travel route database,and proposed a frequent route sequence pattern mining algorithm which can generate mass candidate routes according to the tourists’ type.Finally,a multi-dimensional route searching and ranking mechanism was designed to recommend personalization travel routes.The experimental results illustrate that the proposed method can adopt various factors,i.e.the number of tourist’s travel days,user types and attractions preferences,to help tourists to design personalized travel routes and effectively improve the tour experience.

Key words: Frequent sequence pattern mining, Knowledge graph, Travel route recommendation, User generated data