Personalized recommendation method for scenic spots

Abstract

The invention discloses a personalized recommendation method for scenic spots. The method comprises the following steps of: firstly, constructing a user knowledge graph and a scenic spot knowledge graph by utilizing an original data set; performing feature learning on the user knowledge graph to obtain a first user representation vector and a first project representation vector; performing featurelearning on the scenic spot knowledge graph based on the first user representation vector to obtain a second user representation vector and a second project representation vector; combining the firstuser representation vector and the second user representation vector into a final user representation vector; directly taking the second project representation vector as a final project representation vector; and finally, performing deep interaction on the final user representation vector and the final project representation vector to predict the preference probability of users to the scenic spots, thereby finishing personalized recommendation of the scenic spots. Feature learning of a single knowledge graph can be avoided, so that recommendation accuracy is improved.

Description

technical field [0001] The invention relates to the technical field of personalized recommendation, in particular to a personalized recommendation method for tourist attractions. Background technique [0002] With the development of information technology and the Internet industry, especially the rise of electronic payment, users choose more and more platforms, and travel, travel booking methods and consumption methods are also changing. After service-oriented industries embrace digitalization and smart technology, major platforms have launched smart solutions to promote the digital upgrade of the industry. The data volume of major industries has grown rapidly, and information overload has become a challenge for people to process information. For specific users, how to quickly and accurately locate the content they need in the exponentially growing resources is a very important and extremely challenging matter. For service providers, how to present appropriate products to u...

AI Technical Summary

Problems solved by technology

[0004] Although the personalized recommendation algorithm has been developed for more than ten years, researchers are still working on exploring more efficient recommendation algorithms, but the current personalized recommendation technology still has some imperfections. From the perspective of users, Consider user behavior information from multiple perspectives, refine product categories, focus more on timeliness in recommendation results, higher quality recommendation results, more diverse recommendation content, and more accurate prediction results.

Images