|
Abstract
|
The topic of influence maximization using community-based higher-order optimal percolation in complex networks is highly necessary and important in today's world. Complex networks are part of daily life. They include social media like Facebook and Twitter, biological systems, and transportation grids. Influence spread happens everywhere. For example, in marketing, companies want to reach many customers with low cost. In public health, stopping virus or rumor spread saves lives. Traditional methods use simple pairwise models. They ignore group interactions and communities. But real networks have groups. People influence each other in teams or families, not just one-to-one [3], [6]. Higher-order models fix this. They use hypergraphs to show group effects. This makes predictions better. Communities speed up local spread. Seeds in tight groups reach faster [5]. Optimal percolation finds key nodes efficiently [4]. Combining them gives stronger tools. This research is important for real applications. In social media, ads or news spread via groups and communities. Better methods help fair campaigns or fight misinformation. In epidemiology, higher-order models predict outbreaks more accurately. Group contacts like meetings matter more than pairs. Community-based percolation can find super-spreaders fast. In business, viral marketing needs minimal seeds for max reach. Old greedy methods are slow for big data [1]. New hybrid can scale to billions of nodes. Also, networks change over time. Dynamic higher-order percolation handles that [32]. Importance grows with data explosion. Internet of Things and online platforms create huge complex nets.
Necessity comes from gaps in current work. Most studies use simple graphs [1]. They miss higher-order effects that change dynamics a lot [3]. Community methods improve speed but rarely add groups [5]. Percolation is strong but not community aware [4]. Full integration is rare. This causes wrong results in real cases. For example, ignoring
|