Stage-Aware Preferential Attachment in Resource-Constrained Communities

Jia Yu et al. show community sports organizations (CSOs) evolve through stages with distinct stimuli-response patterns. Yet, growth models rarely condition their attachment mechanisms on developmental stage. Building on WBPA (Topîrceanu et al.) and the staged evolution evidence in CSOs, this project specifies an attachment function that is stage-dependent: early growth favors degree (visibility), mid-stage emphasizes betweenness (brokerage for scale), and mature stages privilege role/competence-based signals (e.g., organizers, resource holders). We would implement this in an agent-based simulator augmented with LLM-agents to generate more realistic behavior and content (Ferraro et al.), then compare to observed longitudinal traces in CSOs. To forecast and diagnose transitions, we’d train a GCNN-style predictor (Ou et al.) that incorporates evolving topology as prior structure, and use separable temporal ERGMs to test whether stage-specific motifs (reciprocity, transitivity, homophily) shift as predicted (cf. Ho et al. on longitudinal collaboration networks). The novelty is the explicit stage-conditioning of the attachment kernel, offering a mechanism-level account for empirical findings like weak densification and assortativity variation across stages. Practically, it yields stage-aligned interventions (e.g., boosting brokers mid-stage vs. investing in role specialization later) and theoretically it reconciles why one-size-fits-all preferential attachment fails to capture multi-stage community growth.

References:

1. Weighted Betweenness Preferential Attachment: A New Mechanism Explaining Social Network Formation and Evolution. Alexandru Topîrceanu, M. Udrescu, R. Marculescu (2018). Scientific Reports.
2. Fostering interdisciplinary collaboration: A longitudinal social network analysis of the NIH mHealth Training Institutes. Eric Ho, Minjeong Jeon, Minho Lee, Jinwen Luo, A. Pfammatter, V. Shetty, Bonnie Spring (2021). Journal of Clinical and Translational Science.
3. Community Sports Organization Development From a Social Network Evolution Perspective— Structures, Stages, and Stimulus. Jia Yu, Meng-jiao Ding, Qingran Wang, Weiqiang Sun, W. Hu (2023). IEEE Transactions on Computational Social Systems.
4. Agent-Based Modelling Meets Generative AI in Social Network Simulations. Antonino Ferraro, Antonio Galli, Valerio La Gatta, Marco Postiglione, Gian Marco Orlando, Diego Russo, Giuseppe Riccio, Antonio Romano, Vincenzo Moscato (2024). International Conference on Advances in Social Networks Analysis and Mining.
5. Opinion Formation Forecasts in Social Networks: A Graph Convolutional Neural Network Approach. Lizhen Ou, Yiping Yao, Wenjie Tang, Haozhe Yuan, Li-li Chen (2023). IEEE International Symposium on Distributed Simulation and Real-Time Applications.