CCChun Cheng

Chun Cheng

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Chun Cheng has a total of 18 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Phase Transitions and Non-Mainstream Norms in a Coevolution Model of Opinions and Actions with Three-Way Decision Thresholds

    by Zhengzheng Jiang, Naibo Wang, Enze Li, Chun Cheng

    In social systems, the coevolution of individual opinions and external actions is fundamental to understanding collective phenomena such as group decision-making and innovation diffusion. Traditional models often simplify the opinion–action relationship into linear or probabilistic mappings, which struggle to capture the widely observed opinion–action inconsistency and nonlinear phase transitions. From the perspective of social physics, this paper integrates the threshold concept of three-way decision theory to construct an opinion–action coevolution model on a dual-layer network. The model introduces decision thresholds that divide the opinion to action process into acceptance, delay, and rejection regions, thereby quantifying decision uncertainty and providing a structured representation of the intermediate hesitation state. Through large-scale agentbased simulations, we explore the effects of mass media, innovation leaders, and the strength of opinion–action coupling on the emergence of collective social innovation. Our findings reveal three main results: (1) Media and innovation leaders exert a synergistic driving effect on innovation diffusion; (2) Excessive opinion–action coupling can trigger a negative feedback trap, whereby strong social conformity pressure suppresses rather than accelerates innovation adoption; (3) When leaders exhibit opinion–action inconsistency, the system may evolve into a stable non-mainstream norm characterized by high opinion consensus but persistently low action levels. These results offer a new mechanistic explanation for phenomena such as delayed adoption and social norm lock-in. The robustness of these findings is further confirmed across diverse network topologies, alternative opinion dynamics models, and different update protocols.

  • Open Access

    Modified Hegselmann-Krause Model for Enhancing Opinion Diversity in Social Networks

    by Chun Cheng, Jiahao Gu, Siyan Lu, Weiping Ding

    A goal of opinion dynamics modelers has long been to find a social science based model that generates strong diversity: smooth, stable, possibly multi-modal distributions of opinions. However, the classic Hegselmann-Krause (HK) model, which relies on similarity assumptions, is often criticized for producing only weak diversity. Here, weak diversity refers to the eventual convergence of group into several opinion clusters, where the gap between opinion clusters is limited by the confidence interval ε, and all individuals within the same cluster share a single opinion. To address this limitation, this paper proposes a modified HK model (MHK) that incorporates social conformity and uniqueness theory from social psychology to enhance opinion diversity within the network. Social conformity theory accounts for the discrepancy between individuals’ expressed opinions and their implicit opinions, while uniqueness theory establishes a new lower bound for individuals’ confidence intervals. Our findings demonstrate that social conformity and uniqueness collectively promote strong diversity both within and between clusters. Consequently, our modified model effectively addresses the limitations of the classic HK model, thereby improving its capacity to capture strong diversity in opinion dynamics.

  • ViTH-RFG: Vision Transformer Hashing With Residual Fuzzy Generation for Targeted Attack in Medical Image Retrieval

    by Weiping Ding, Chuansheng Liu, Jiashuang Huang, Chun Cheng, Hengrong Ju

    The rapid advancement of medical technology has led to an exponential increase in the volume of medical images. To optimize clinical practice, physicians often require efficient retrieval of images from various medical image databases. However, the inherent structural and textural similarities between medical images add complexity to retrieval. Traditional methods and deep learning-based hashing techniques have limitations, including their sensitivity to adversarial examples and insufficient robustness. Adversarial learning through Generative Adversarial Networks (GANs) offers an effective solution to address these challenges. In this paper, we propose a Vision Transformer Hashing with Residual Fuzzy Generation (ViTH-RFG) for targeted attacks in medical image retrieval. ViTH-RFG comprises a Vision Transformer Hashing (ViTH) model, a prototype network, a residual fuzzy generator, and a discriminator. ViTH employs the Vision Transformer (ViT) as a feature extraction module and, to better adapt to medical images, cleverly integrates a Spatial Pyramid Pooling (SPP) layer into the Multi-Head Attention (MHA) mechanism to extract and fuse medical image features at different scales. The prototype network takes image labels as input, transmitting the extracted semantic features to the residual fuzzy generator to facilitate the generation of adversarial examples. In the residual fuzzy generator, we introduce residual fuzzy blocks and embed a fuzzy layer to handle the complex relationships between feature maps and the corresponding generated examples, thereby enhancing the model's performance in the face of targeted attacks. The discriminator is responsible for identifying whether an image is an original or a generated example. Finally, experiments on ChestX-ray14 and ISIC 2018 verified the effectiveness of ViTH-RFG in targeted attack scenarios


  • Cooperative evolution with opinion formation in a complex social environment

    by Yun Luo, Yuke Li, Chun Cheng

    Cooperation is a significant aspect in the daily activities of human or animal populations that involves the process of agents making decisions based in part on the ideas or beliefs of cooperators. The importance of communication in the evolutionary game cannot be overstated. In this paper, we offer a co-evolutionary game model on a communication network, including expressed and private opinions. Then, we present our theoretical analysis of scenarios characterized by different parameters, elucidating the relationship between the agents’ opinion formation process and the decision-making process. Finally, we run simulations with our proposed model. Surprisingly, we discover that whereas communication does not increase cooperation on Erdös–Rényi random networks, it does so on the Barabási–Albert scale-free network. Meanwhile, we discover that in the simulation results, the average of private opinions (simplified as values within [0,1]) changes in the same direction of the percentage of cooperators. Furthermore, we discover that deceivers who conceal their true private opinions may, to some extent, foster the formation of collaboration in the human population, which goes against our common sense.