Community-Driven Shortest Path Optimization Framework for Vaccine Delivery in Social Community Networks

The Shortest Path (SP) problem is applicable to a wide range of applications pertaining to knowledge propagation, weighted social networks, and others. Conventional shortest path algorithms, such as Dijkstra’s and Bellman–Ford’s, are ineffective in a weighted social network, or community-driven networks, where relationships and interactions are often dynamic, leading to continuously evolving network structures. Additionally, in such networks, the significance of specific nodes or communities may change over time, further complicating the task of identifying the shortest path. Motivated by the limitation, Clique Percolation and Monte Carlo-based Shortest Path (CP-MCBSP) framework has been proposed, applicable to vaccine delivery network model. The proposed CP-MCBSP algorithm integrates clique percolation and Monte Carlo techniques for identifying communities and shortest path computation. Clique sizes are adjusted for community exploration, and Monte Carlo simulations are used for the best route selection. The algorithm’s uniqueness stems from its combination of community-based knowledge and unpredictable assessments, which enables effective navigation in complex networks with overlapping communities. The optimal choice of paths determined by Monte Carlo improves durability in dynamic conditions in networks. In comparison to current shortest path algorithms, CP-MCBSP offers a 7.04% boost in computational efficiency, and a 5.28% reduction in approximation error. The integrated difficulties of community discovery and shortest possible route compositions were employed in road network datasets from California, San Francisco, and West Bengal. The findings show that the proposed technique outperforms previous strategies in this particular area.