Post-collisional potassic-ultrapotassic volcanic rocks are widely developed in the Lhasa block, Qinghai-Tibet region. In this paper, it was observed for the first time that dilute alkali elements–Li, Rb, and Cs–exhibit supernormal enrichment in the research data of numerous potassic-ultrapotassic volcanic rocks in the Lhasa block. At the same time, there are few studies on the genesis of enrichment. Currently, 568 sets of published volcanic rock data and 8 sets of measured data in the Lhasa block are sorted in detail, and the genesis of dilute alkali element enrichment is explained by means of geochemical research methods. It is believed that the high degree of magmatic fractionation of potassic-ultrapotassic volcanic rocks in the Lhasa block is the main reason for the abnormal enrichment of dilute alkali elements such as Li, Rb, and Cs. The abnormal enrichment area is mainly located in the central and western parts of the Lhasa block with an age range of 25–13 Ma. The discrimination range of Zr/Hf and Nb/Ta with a high degree of differentiation of potassic-ultrapotassic volcanic rocks in the Lhasa block is divided by analogy with the research results of highly fractionated granites, which provides a reference for the study of the supernormal enrichment of dilute alkali elements in potassic-ultrapotassic volcanic rocks in this area.

Many community detection algorithms suffer from problems such as low accuracy and efficiency, overfitting or underfitting, and high time complexity. To address the issues in global optimization algorithms proposed so far, GOMK-BP, a community detection algorithm based on the probabilistic graph model, is proposed in this paper. It takes the fully connected Markov random field (MRF) as the probabilistic graph model, maps the node’s similarity and importance indexes to the energy function, which is then optimized to improve the performance of the model, and the combination of weighted sum and exponential function makes the algorithm achieve a better fit. To improve the efficiency and accuracy, the traditional message propagation algorithm is improved, and the convergence is accelerated by using compressed sensing. Because of the fully connected MRF may be affected by the high computational complexity, GOMK-BP adjusts the model structure, and introduces sparse connections to balance the computational efficiency and model performance. Finally, the community structure is obtained. The proposed algorithm is tested on some artificial networks and real networks, the results show that GOMK-BP can detect high-quality community structures.

A special kind of complex networks known as multiplex networks occurs when a single collection of entities engages in a variety of interactions. Moreover, in a multiplex network, the topologies of other levels may have an impact on how links form in one layer. At present, a lot of solutions to the link prediction problem have been developed. However, few of them have tackled link prediction in multiplex networks using graph neural networks. In this paper, we propose a new approach to conduct link prediction in multiplex networks using graph neural networks. The proposed approach first combines all layers of a network to form an information aggregation graph, and then reconstruct an information filter graph by leveraging the similarities of nodes. Next, for the target node pair, we extract its enclosing subgraphs from the target layer, auxiliary layers, and information filtering graph, and then predict the connection of the target node pair by inputting the subgraphs with feature matrices into a multilayer perceptron. The proposed solution offers better performance for link prediction in multiplex networks, as shown by the experimental results on six real networks.