Knowledge Graph based Retrieval-Augmented Generation (KG-RAG) technology is an effective method to improve the performance of Large Language Models (LLMs) in medical question answering tasks. However, there is an inherent difference between structured knowledge graphs and sequential LLMs. The existing method of converting triples of knowledge graph into text through predefined templates leads to loss of structural information and excessive context length, resulting in alignment difficulties between the two. In this paper, we propose a GNN-LLM Aligned Medical knowledge QuestionAnswering model (GLAM-QA). The model first pretrains a GNN to extract structural information from the knowledge graph, realizes the interactive fusion of graph representations and LLM vocabulary embeddings through a cross attention layer, and then generates graph tokens via a projector for input to the LLM. It can retain graph structural information without finetuning the LLM and provide accurate and concise answers. Experiments on three public datasets show that compared with five comparative algorithms, the proposed GLAM-QA achieves better performance in BertScore, F1-Score, and BLEU-1 metrics. This research provides an effective solution to the alignment problem between knowledge graphs and LLMs, and improves the accuracy and practicality of medical question answering systems.

With the rapid development of high-throughput sequencing technologies, multi-omics integration analysis has become a core means to decipher complex disease mechanisms, and graph neural networks (GNNs) have shown significant advantages in multi-omics data fusion due to their strong nonlinear modeling and relational reasoning capabilities. However, such models typically rely on large-scale labeled data for training, while small-sample disease scenarios are prevalent in biomedical research. The limited sample size in these scenarios does not support stable deep model training, severely restricting the application of precision medicine in critical disease fields. Directly constructing GNN models under small-sample settings leads to severe overfitting and training instability, resulting in a significant decline in model generalization performance. In this paper, we propose a Cross-disease knowledge transfer framework for Small-sample Omics analysis with GNNs (CSOG), through a pretraining and parameter-freezing fine-tuning paradigm. To the best of our knowledge, we are the first to explore the GNN cross-disease knowledge transfer framework for small-sample omics analysis. Experimental evaluations on three independent small-sample datasets show that compared with 10 state-of-theart baseline methods, the proposed method achieves significantly improved average classification accuracy and maintains stability across different sample sizes.

Graph computing in cloud-edge collaborative environments faces critical challenges in distributed task processing, particularly in fundamental operations such as subgraph isomorphism that underpins triangle counting applications. In typical architectures where data streams are transmitted from edge collectors to cloud masters, conventional approaches employ reservoir sampling to distribute edge streams among workers for triangle estimation. However, the computational accuracy degradation is caused by cross-domain edge distribution strategies. In this paper, we propose a cloud-edge collaborative framework for distributed triangle counting. We employ spectral clustering analysis to reveal latent domain relationships that guide edges distribution. Our experimental evaluation uses streaming data with global relative error measurement across multiple datasets, demonstrating superior performance over existing algorithms.

As social networks continue to expand, an increasing number of people prefer to use social networks to post their comments and express their feelings, and as a result, the information contained in social networks has grown explosively. The effective extraction of valuable information from social networks has attracted the attention of many researchers. It can mine hidden information from social networks and promote the development of social network structures. At present, many ranking node approaches, such as structural hole spanners and opinion leaders, are widely adopted to extract valuable information and knowledge. However, approaches for analyzing edge influences are seldom considered. In this study, we proposed an edge PageRank to mine shortcuts (these edges without direct mutual friends) that are located among communities and play an important role in the spread of public opinion. We first used a network-embedding algorithm to order the spanners and determine the direction of every edge. Then, we transferred the graphs of social networks into edge graphs according to the ordering. We considered the nodes and edges of the graphs of the social networks as edges and nodes of the edge graphs, respectively. Finally, we improved the PageRank algorithm on the edge graph to obtained the edge ranking and extracted the shortcuts of social networks. The experimental results for five different sizes of social networks, such as email, YouTube, DBLP-L, DBLP-M, and DBLP-S, verify whether the inferred shortcut is indeed more useful for information dissemination, and the utility of three sets of edges inferred by different methods is compared, namely, the edge inferred by ER, the edge inferred by the Jaccard index. The ER approach improves by approximately 10%, 9.9%, and 8.3% on DBLP, YouTube, and Orkut. Our method is more effective than the edge ranked by the Jaccard index.