Personalized recommendation method and system based on collaborative filtering and deep learning

Abstract

The invention provides a personalized recommendation method and system based on collaborative filtering and deep learning, and the method comprises the steps: obtaining historical behavior feature data of commodities purchased by a user, carrying out the preprocessing, and sorting the purchasing behaviors of the user according to the time, wherein the sorted data is called a behavior feature sequence of the user; performing the personalized recommendation system modeling, which comprises the steps of obtaining input vectors of a user and a commodity from an interaction matrix, then respectively generating embedded vectors of the user and the commodity, weighting the embedded vectors through an attention neural network, and performing linear and nonlinear interaction on the weighted embedded vectors, thereby obtaining the explicit and implicit relationship between the user and the commodity; finally, estimating the click rate of the user to the commodity; and training and testing the model by using the user behavior characteristic sequence. According to the method, the collaborative signals of the users and the commodities are fully mined, a basis is provided for capturing personalized demands of the users, and the accuracy and interpretability of a recommendation system can be improved.

Description

technical field [0001] The invention relates to the technical field of recommendation systems in the Internet, in particular to a personalized recommendation method and system based on collaborative filtering and deep learning. Background technique [0002] With the networking of human life, the amount of data in the network is increasing explosively, and the information contained in these data is also increasing day by day. In people's daily network life, more and more applications begin to pay attention to using these Information to improve the user's online experience, the recommendation system came into being. Corresponding technologies are also increasing, such as: [0003] CN109410001B provides a kind of commodity recommendation method, system, electronic equipment and storage medium, obtains the weight value of the text phrase of commodity, and carries out vectorization to text phrase and obtains corresponding weighted text word vector; The weighted text word vector...

AI Technical Summary

Problems solved by technology

[0010] 1. The currently widely used collaborative filtering model adopts an embedded method when calculating the hidden feature vectors of users and products. In this method, most models represent users and products as one-hot encoding, and then use the embedding matrix To get the corresponding hidden vector, this method does not make full use of the information in the user-product interaction matrix, for example, which products the user has purchased in the past, and which users have purchased a certain product in the past, these cannot be known through this coding method, and This information is often very important when making personalized recommendations to users

[0011] 2. Recommendation models based on deep learning aim to mine the nonlinear feature relationship between users and items, but they often ignore the synergy signals explicitly encoded in low-order interactions, and these linear feature relationships captured in low-order interactions are often The key to personalized recommendations

[0012] 3. When using the user’s past purchase information to recommend products to users, not all products have the same impact on the user’s next purchase behavior. For example, users who prefer electronic products have bought mice, keyboards and hats in the past, then , the system should pay more attention to the impact of the mouse and keyboard on the next purchase behavior, but in many recommendation systems, this idea has not been reflected

Images