Multi-context information fused personalized place recommendation method and device

Abstract

The invention provides a multi-context information fused personalized place recommendation method and device. The method includes the following steps: 1, obtaining geographical label photograph data;2, mining geographical label photograph multi-context information, evaluating context information such as the place popularity, the place popularity fluctuation, a spatial distance, user-based collaborative filtering and the user comment based text similarity, and calculating corresponding scores thereof; 3, constructing a "user-place" correlation feature vector on the basis of transforming the personalized place recommendation problem by using a ranking learning based personalized place recommendation model, and using the ranking learning method to establish a place recommendation model; and4, selecting the top n personalized places in prediction values according to the model in the step 3, and generating a recommendation result and evaluating the recommendation result. The method and device can effectively improve the accuracy and recall rate of personalized place recommendation.

Description

technical field [0001] The present invention relates to the technical field of user recommendation systems, in particular to a personalized place recommendation method that combines multiple context information. Background technique [0002] Personalized location recommendation can provide people with good location-based services. Existing methods have achieved certain recommendation effects in location recommendation, but there are still some shortcomings: First, the collaborative algorithm only uses the user's check-in location However, due to the sparsity of the user's check-in location, the search of adjacent users is not accurate enough, and the recommendation accuracy is poor; secondly, when calculating the probability of a user's visit to a candidate location under the influence of spatial distance in the existing theory, all All the check-in location information of the user does not take into account the individual factor of the travel distance difference of differen...

AI Technical Summary

Problems solved by technology

[0003] The technical problem to be solved by the present invention is to provide a personalized location recommendation method and equipment that integrates multiple contextual information to solve the above-mentioned technical defects in view of the above-mentioned current existing personalized location recommendation methods that have poor accuracy and do not consider user differences.

Images