In this paper, an intelligent multiple Quality of Service (QoS) constrained traffic path allocation scheme with corresponding algorithm is proposed. The proposed method modifies deep Q-learning network (DQN) by graph neural network (GNN) and prioritized experience replay to fit the heterogeneous network, which is applied for production management and edge intelligent applications of smart factory. Moreover, through designing the reward function, the learning efficiency of the agent is improved under the sparse reward condition, and the multi-object optimization is realized. The simulation results show that the proposed method has high learning efficiency, and strong generalization ability adapting the changing of topological structure of network caused by network error, which is more suitable than the compared methods. In addition, it is also verified that combining the field knowledge and deep reinforcement learning (DRL) can improve the performance of the agent. The proposed method can achieve good performance in the network slicing scenario as well.

People’s needs are constantly changing, and the performance of traditional recommendation algorithms is no longer enough to meet the demand. Considering that users’ preferences change with time, the users’ behavior sequence hides the evolution and change law of users’ preferences, so mining the dependence of the users’ behavior sequence is extremely important to predict users’ dynamic preferences. From the perspective of constructing users’ dynamic preferences, this paper proposes a users’ dynamic preference model based on users’ behavior sequences. Firstly, the user’s interest model is divided into short-term and long-term interest models. The short-term interest reflects the user’s current preference, and the long-term interest refers to the user’s interest from all his historical behaviors, representing the user’s consistent and stable preference. Users’ dynamic preference is obtained by integrating short-term interest and long-term interest, which solves the problem that the user’s preference cannot reflect the change in the user’s interest in real-time. We use the public Amazon review dataset to test the model we propose in the paper. Our model achieves the best performance, with a maximum performance improvement of 15.21% compared with the basic model (BPR, NCF) and 2.04% compared with the sequence model (GRU4REC, Caser, etc.), which proves that the user’s dynamic preference model can effectively predict the user’s dynamic preference. Users’ dynamic preferences are helpful in predicting users’ real-time preferences, especially in the field of recommendation.