1 August 2021 in Pattern Recognition Letters
18 January 2022 in Frontiers in Physics
Inspired by lots of applications like viral marketing of products and transmitting information in a network, ranking the spreading ability of nodes in the network has been widely studied. At present, the above problem is mostly studied on unsigned networks which only contain positive relationships (e.g., friend or trust) between users. In real-world networks, there usually exist both positive relationships and negative relationships (e.g., foe or distrust) between users. Based on this, we aim to find the influential spreaders in a signed network which meet the requirement of real scene. Moreover, when the spreading only aims to affect a specific group of nodes instead of all nodes, such as promoting cigarette, a new problem called localized targets spreading problem was come up with. Localized targets spreading problem has been studied on unsigned networks, but it is still open for signed networks. Thus, in this paper, we propose a new method, called local influence matrix (LIM) method, which aims to find the seed nodes set with maximum positive influence on a specific group of targets but with minimum influence on the non-target nodes in signed social networks. Simulation results show that our method performs well on real networks.
10 March 2026 in Medical Image Analysis
1 March 2026 in IEEE Transactions on Computational Biology and Bioinformatics
Alzheimer's disease (AD) is a progressive neurodegenerative disorder that remains a global challenge due to its complex pathology and the lack of definitive diagnostic tools. This paper introduces an innovative approach to predicting and analyzing Alzheimer's disease by constructing signed brain network models and leveraging signed graph neural network technologies. By modeling the brain network as a signed graph that incorporates both positive and negative correlations, we capture the nuanced interactions between brain regions more effectively than traditional methods. We utilize graph convolutional networks (GCNs) and their variants to process these signed brain networks, significantly improving the accuracy of Alzheimer's disease prediction. Comparative analysis reveals that the signed graph model outperforms its unsigned counterparts in diagnostic precision (with an improvement of at least $19\%$), emphasizing the importance of incorporating negative correlations in neural interactions. Furthermore, precisely because of the additional negative edge information that we can utilize both positive and negative attention matrices, derived from these prediction tasks, to determine important brain region biomarkers. This work is an attempt to systematically validate the role of negative information through comparisons of different signed graph variants, which holds particular promise for enhancing Alzheimer's disease diagnostic accuracy at early stages. We believe that this approach will have significant clinical applications in the future.
9 July 2024 in Physica D: Nonlinear Phenomena
15 April 2025 in Physica Scripta
As a fundamental dynamical process in real-world systems, spreading is extensively studied in various fields. Several spreading models are proposed to simulate some real scenes, such as information propagation or epidemic contagion. One of the key challenges in the study is known as influence maximization problem, whose goal is to find k nodes that have the greatest influence on all the nodes in the network. The influence maximization problem has been proven to be NP-hard, leaving significant challenges in finding accurate solutions. Substantial effort has been dedicated to addressing this problem. More recently, in cases where spreading only aims to influence a specific node set rather than all the nodes, like promoting cigarette advertisements whose potential customers are only adults, a new problem called localized targets influence maximization problem came up. In this paper, we propose a novel method to address this localized target influence maximization problem based on an influential score discount method, which not only considers the influence of candidate nodes on the target node set, but also takes into account the overall influence of the candidate node set on the target node set. The simulation results show that our method can effectively promote infecting target nodes.