Recommendation method, system and equipment based on collaborative filtering and medium

Abstract

The invention discloses a recommendation method, system and device based on collaborative filtering and a medium, and the method comprises the steps: obtaining the historical score data of a user for an item, calculating the user similarity on the historical score data according to a new similarity method, obtaining an optimal parameter k through the new method, and calculating the nearest neighbor of the optimal user, and removing noise users in the nearest neighbor by using an optimized local density outlier algorithm, and finally calculating a recommendation result by using the denoised user nearest neighbor and an average weighted prediction formula to realize innovative recommendation application. The advantages of outlier detection and particle swarm optimization are combined with the collaborative filtering recommendation algorithm, so that the collaborative filtering recommendation system is effectively improved, the time complexity of the system is reduced, and the problems of data sparsity in the collaborative filtering recommendation algorithm and poor recommendation effect caused by a single improvement method are further solved. And the precision is low.

Images