Xiaolong Deng

To address the issue of traditional defense strategies being inflexible in handling new types of attacks due to resource rigidification, this study focuses on constructing a self-generating defense strategy model through algorithm design. This model allows for flexible adjustments to dynamically adapt to the specific equipment’s network topology, using simulation design to construct a network attack and defense environment. This addresses the shortcomings of existing simulation designs that lack real-world scenario features and practical network attack analysis, thereby providing an experimental platform for deep research on defense strategies. Through theoretical analysis of deep reinforcement learning algorithms, a defense strategy generation scheme based on these algorithms is designed and optimized according to environmental characteristics. Finally, comparative experimental simulations are conducted by evaluating win rates and resource consumption. This study proposes a self-generating scheme of strategies based on deep reinforcement learning within the new network equipment architectures. Our scheme automatically generates defense strategies to counter emerging network attacks in the network, flexibly allocates network resources, executes dynamic proactive defense strategies, and automatically eliminates the impact of unknown attacks. In addition, it continuously expands the expert knowledge base and uses experimental data from parallel network environments to enhance the defensive capabilities of intelligent agents.

In the current era characterized by exponential data growth, industries across the spectrum have amassed enormous textual data repositories brimming with untapped value. The challenge lies in efficiently and precisely mining this information, posing a universal dilemma for academia and industry alike. Introducing the MBG-C Ner model, this study innovatively fuses MacBERT, BiGRU, and CRF methodologies to adeptly grasp contextual nuances, tackle long-range dependencies, and refine label sequence decoding, thereby enhancing named entity recognition capabilities. To validate its efficacy, rigorous experiments spanning diverse datasets—MSRA, OntoNotes-4, Weibo NER, and Resume—were undertaken. The outcomes underscore MBG-C Ner's remarkable superiority over cutting-edge techniques in Chinese named entity recognition, evidenced by peak F1 scores of 95.70%, 80.17%, 69.76%, and 97.46%, respectively. This milestone not only validates MBG-C Ner's exceptional performance but also opens new avenues and strategies for advancing Chinese named entity recognition research.

The study on the graduate students’ comprehensive quality evaluation system based on the “Five Educations in Parallel” concept aims to construct a comprehensive, scientific, and operational evaluation...

Link prediction is critical to completing the missing links in a network or to predicting the generation of new links according to current network structure information, which is vital for analyzing the evolution of a network, such as the logical architecture construction of MEC (mobile edge computing) routing links of a 5G/6G access network. Link prediction can provide throughput guidance for MEC and select appropriate c nodes through the MEC routing links of 5G/6G access networks. Traditional link prediction algorithms are always based on node similarity, which needs predefined similarity functions, is highly hypothetical and can only be applied to specific network structures without generality. To solve this problem, this paper proposes a new efficient link prediction algorithm PLAS (predicting links by analysis subgraph) and its GNN (graph neural network) version PLGAT (predicting links by graph attention networks) based on the target node pair subgraph. In order to automatically learn the graph structure characteristics, the algorithm first extracts the h-hop subgraph of the target node pair, and then predicts whether the target node pair will be linked according to the subgraph. Experiments on eleven real datasets show that our proposed link prediction algorithm is suitable for various network structures and is superior to other link prediction algorithms, especially in some 5G MEC Access networks datasets with higher AUC (area under curve) values.