Personalized scenic spot recommendation method and system based on positive and negative feedback portrait coding of user

Abstract

The invention provides a personalized scenic spot recommendation method based on positive and negative feedback portrait coding of a user, which comprises the following steps: collecting historical evaluation information of the scenic spot and tourism attribute data of the scenic spot; obtaining positive and negative evaluation attractions of the user; transforming all scenic spots and their tourism attribute values into triples to construct scenic spots knowledge map; training the entity, attribute and attribute value of the scenic spot in the triple by the scoring function, so that the entity, attribute and attribute value of the scenic spot are converted into the vector representation form; obtaining positive and negative feedback portrait coding of scenic spots by users; calculating the similarity between the positive feedback image coding and the scenic spot entity vector to get the scenic spot that the user likes. Then the negative feedback image coding is used to optimize the ranking. The invention encodes the negative feedback portrait of the scenic spot characteristics by the user, adds the similarity degree of the scenic spots that the user hates on the basis of the favorite interest list, calculates and optimizes to obtain the final recommendation list, and provides the more accurate recommendation of the scenic spots for the user.

Description

technical field [0001] The present invention relates to technical fields such as machine learning, knowledge graph, and intelligent recommendation, and in particular to a personalized scenic spot recommendation method and system based on user positive and negative feedback image coding. Background technique [0002] With the changes in people's living standards and lifestyles, more and more people need to go out for activities, such as watching movies, dining and traveling, and people's demand for personalized recommendation of location information is becoming greater and greater. It is particularly important to recommend suitable scenic spots at this time. Therefore, how to construct a user portrait based on the evaluation feedback information of the user's visited attractions is the key to a personalized attraction recommendation system. Most of the traditional location-based recommendations are based on the popularity statistics of scenic spots, which cannot meet the per...

AI Technical Summary

Problems solved by technology

[0003] In view of the shortcomings of the prior art described above, the purpose of the present invention is to provide a personalized scenic spot recommendation method and system based on user positive and negative feedback image coding to solve the problem of low accuracy of scenic spot recommendation in the position-based recommendation method in the prior art. Personalization of recommendation results and low level of scenic spot experience

Images