Recommendation method and system based on deep collaborative filtering

Abstract

The invention provides a recommendation method and system based on deep collaborative filtering, and the method comprises the steps: obtaining massive user and project score data from the Internet, and preprocessing the user and project score data; performing embedding operation on the preprocessed data, and completing the linear representation process of the interaction function; sending the complete linear representation of the interaction function into a deep neural network to obtain a complete adaptive interaction function; predicting scores, and generating a recommendation list. In the same process, the Hadamard product and linear feature extraction process is carried out on the implicit vectors of the user and the item, and a more complex interaction function can be learned to a certain extent by adding a hidden layer on the basis of combining linear representation of the user and the item, so the performance of the model is improved, and the user experience is improved.

Description

technical field [0001] The present invention relates to the technical field of computer technology collaborative filtering recommendation, in particular to a recommendation method and system based on depth collaborative filtering. Background technique [0002] In the era of information explosion, the problem of information overload has become one of the difficulties faced by everyone. As a core information push method, the recommendation system can timely and efficiently provide information that online users are interested in, save users' time cost and system resources, and generate significant market value. [0003] The traditional recommendation system selects some popular products from a large number of products and recommends them to users, but not every user likes popular products. With the development of the times, users' personalized characteristics are becoming more and more obvious, and attention should be paid to user preferences and interests. Featured personaliz...

AI Technical Summary

Problems solved by technology

[0005] However, the traditional matrix factorization algorithm maps the learned user latent vector and item latent vector into the same space, and then the linear dot product operation constrains the representation ability of the interaction function, that is, it is difficult to learn the complex nonlinear relationship between users and items. relationship, which greatly limits the accuracy of the final generated recommendation list

Images