Study on Time-sharing Reservation Recommendation System of Scenic Spots Based on Relationship Graph

Abstract

With the continuous development of the Internet, the content and information in the network are increasing explosively. Nowadays, with the rapid increase of data volume, many Internet companies need to consider how to promote their products to potential users and realize personalized recommendation. To solve this problem, people need to find relationships between data to build a relationship graph. Then, the potential relationship between each node and other nodes can be found according to the relationship graph, and personalized recommendation can be finally realized. Based on the reservation records of scenic spots in Gansu Province from 2020 to 2021, this paper builds a relationship graph between tourists and scenic spots. Neo4j graph database is used to store the relationship graph and display it intuitively. Then the relationship graph is used as the input of the recommendation system model in this paper to obtain the embedding representation of each type of tourists. Then, K-means is used to cluster tourists to obtain the reserved scenic spots of each category and calculate the popularity of scenic spots, so as to obtain TOP-K scenic spot recommendation and realize personalized recommendation.