A method and device for recommending scenic spots based on tourism knowledge graph

CN112667877BActive Publication Date: 2025-02-28SHAANXI NORMAL UNIV
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202011567776.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2020-12-25
Publication Date
2025-02-28
Estimated Expiration
2040-12-25

AI Technical Summary

Technical Problem

The existing tourism recommendation system has shortcomings in project cold start and data sparseness, and cannot provide users with personalized attractions recommendations.

Method used

By crawling the data of the tourism website, a tourism knowledge graph is constructed, and a knowledge graph representation learning algorithm is used to embed entities and relationships into low-dimensional vectors to calculate the similarity between attractions. Combining the user-attraction rating matrix, combining similarity calculations, we finally generate a personalized attraction recommendation list.

Benefits of technology

It effectively solves the problems of project cold start and data sparsity in traditional recommendation systems, and improves the accuracy and personalization of recommendations.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN112667877B_ABST
    Figure CN112667877B_ABST
Patent Text Reader

Abstract

A method and device for recommending scenic spots based on a tourism knowledge graph, the recommendation method includes the following steps: crawling tourism website data information and cleaning and arranging it to construct a tourism field knowledge graph; embedding entities and relationships of the tourism field knowledge graph into low-dimensional vectors, and calculating the similarity between scenic spots; constructing a user-scenic spot rating matrix, and calculating the similarity between scenic spots according to the user-scenic spot rating matrix; fusing the similarity calculated according to knowledge representation learning with the similarity calculated by the user-scenic spot rating matrix to obtain the final similarity; according to the final similarity, the top K entities with the highest similarity to the target entity are obtained to form a recommendation list. The present invention effectively solves the problems of project cold start and data sparsity in traditional recommendation systems, and can more accurately make personalized recommendations for target tasks.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Knowledge atlas-based personalized tourist route planning method

    CN107679661A

  • Personalized tourist attraction recommending method based on knowledge domains map

    CN107729444A