Deep reinforcement learning-based recommendation method and system with negative feedback

Abstract

The invention discloses a deep reinforcement learning-based recommendation method and system with negative feedback, and the method comprises the steps: collecting commodity feature information, and collecting user behavior data to obtain positive and negative feedback behavior vectors of a user; performing feature extraction on the positive and negative feedback behavior vectors of the user through a feature extraction network model to obtain a positive and negative feedback feature mixed state vector of the user; training a deep deterministic strategy gradient model composed of a strategy network and an estimation network by using the positive and negative feedback feature mixed state vector of the user until the model converges; and generating a positive and negative feedback feature mixed state vector according to historical behaviors of a user needing work recommendation, and generating a user recommendation commodity list for the user for selection through the trained deep deterministic strategy gradient model to complete the user work recommendation. The parameter updating of the related neural network can be delayed, so that the correlation between the networks is reduced,and the training speed and accuracy of the recommendation method are improved.

Description

technical field [0001] The invention belongs to the technical field of data mining and recommendation, and more specifically relates to a recommendation method and system based on deep reinforcement learning with negative feedback. Background technique [0002] In recent years, with the vigorous development of e-commerce, the technology in the field of user recommendation has also developed rapidly. For example, various technologies based on collaborative filtering, machine learning, and deep learning have been applied to recommendation methods. However, traditional recommendation methods have various problems, such as methods based on collaborative filtering, methods based on traditional machine learning, or methods based on deep learning. They all have the characteristics of static recommendation algorithms, that is, they cannot dynamically perform React, and most methods follow a fixed algorithm similar to greed to make recommendations, which will make them pay too much a...

AI Technical Summary

Problems solved by technology

[0005] In view of the above defects or improvement needs of the prior art, the present invention provides a recommendation method and system based on deep reinforcement learning with negative feedback, aiming at solving the problems of too slow learning rate and low accuracy of existing user recommendation methods

Images