Personalized Tourism Route Recommendation System Based on Dynamic Clustering of User Groups

Tourism path dynamic planning is an asynchronous group model planning problem. It is required to find group patterns with similar trajectory behavior under the constraint of unequal time intervals. Traditional trajectory group pattern mining algorithms often deal with GPS data with fixed time interval sampling constraints, so they can not be directly used in coterie pattern mining. At the same time, traditional group pattern mining has the problem of lack of semantic information, which reduces the integrity and accuracy of personalized travel route recommendation. Therefore, this paper proposes a semantic based distance sensitive recommendation strategy. In order to efficiently process large-scale social network trajectory data, this paper uses MapReduce programming model with optimized clustering to mine coterie group patterns. The experimental results show that: under MapReduce programming model, coterie group pattern mining with optimized clustering and semantic information is superior to traditional group mode in personalized travel route recommendation quality, and can effectively process large-scale social network trajectory data.