Dynamic (temporal) graphs are a convenient mathematical abstraction for many practical complex systems including social contacts, business transactions, and computer communications. Community discovery is an extensively used graph analysis kernel with rich literature for static graphs. However, community discovery in a dynamic setting is challenging for two specific reasons. Firstly, the notion of temporal community lacks a widely accepted formalization, and only limited work exists on understanding how communities emerge over time. Secondly, the added temporal dimension along with the sheer size of modern graph data necessitates new scalable algorithms. In this paper, we investigate how communities evolve over time based on several graph metrics under a temporal formalization. We compare six different algorithmic approaches for dynamic community detection for their quality and runtime. We identify that a vertex-centric (local) optimization method works as efficiently as the classical modularity-based methods. To its advantage, such local computation allows for the efficient design of parallel algorithms without incurring a significant parallel overhead. Based on this insight, we design a shared-memory parallel algorithmDyComPar, which demonstrates between 4 and 18 fold speed-up on a multi-core machine with 20 threads, for several real-world and synthetic graphs from different domains.

This paper introduces a novel framework that integrates agent-based mobility simulation with social network analysis to model epidemic diffusion, using Las Vegas as a high-stakes test bed. By mining real-world transportation, demographic, and geographic data, we construct a high-resolution, dynamic contact network that captures the heterogeneous interactions among key subpopulations. Applying graph theory and diffusion modeling to this evolving network demonstrates that epidemic risk is structurally determined by critical transmission hotspots (major resorts, transport hubs) and is highly sensitive to the initial seeding location, with an introduction at the airport leading to a far more explosive outbreak than one in a residential zone. We demonstrate how distinct population groups play unique network roles: tourists act as primary importation vectors, while hospitality workers serve as crucial bridging nodes between otherwise disconnected clusters. The model also shows how some agents achieve “immunity by circumstance” due to mobility patterns that isolate them from high-traffic areas. Our experiments prove that targeted interventions, such as capacity limits at identified hotspots, substantially outperform universal restrictions. The framework validated in Las Vegas provides a transferable tool for strategic epidemic preparedness, offering a new paradigm for data-driven public health policy applicable to any globally connected urban center.

Understanding the structural characteristics and connectivity patterns of large-scale software ecosystems is critical for enhancing software reuse, improving ecosystem resilience, and mitigating security risks. In this paper, we investigate the Maven Central ecosystem, one of the largest repositories of Java libraries, by applying network science techniques to its dependency graph. Leveraging the Goblin framework, we extracted a sample consisting of the top 5,000 highly connected artifacts based on their degree centrality and then performed breadth-first search (BFS) expansion from each selected artifact as a seed node, traversing the graph outward to capture all libraries and releases reachable those seed nodes. This sampling strategy captured the immediate structural context surrounding these libraries resulted in a curated graph comprising of 1.3 million nodes and 20.9 million edges. We conducted a comprehensive analysis of this graph, computing degree distributions, betweenness centrality, PageRank centrality, and connected components graph-theoretic metrics. Our results reveal that Maven Central exhibits a highly interconnected, scale-free, and small-world topology, characterized by a small number of infrastructural hubs that support the majority of projects. Further analysis using PageRank and betweenness centrality shows that these hubs predominantly consist of core ecosystem infrastructure, including testing frameworks and general-purpose utility libraries. While these hubs facilitate efficient software reuse and integration, they also pose systemic risks; failures or vulnerabilities affecting these critical nodes can have widespread and cascading impacts throughout the ecosystem.