Total positive influence domination on weighted networks

TL;DR

This paper introduces new approximation algorithms for the total positive influence dominating set problem in weighted networks, with applications in social networks, blockchain, and distributed systems, supported by extensive experiments.

Contribution

It presents two greedy algorithms and a linear programming based approach for the problem, offering novel solutions with practical applications.

Findings

Algorithms perform well on real and synthetic networks

Trade-offs observed between solution quality and computational efficiency

Potential issues identified for each algorithm

Abstract

We are proposing two greedy and a new linear programming based approximation algorithm for the total positive influence dominating set problem in weighted networks. Applications of this problem in weighted settings include finding: a minimum cost set of nodes to broadcast a message in social networks, such that each node has majority of neighbours broadcasting that message; a maximum trusted set in bitcoin network; an optimal set of hosts when running distributed apps etc. Extensive experiments on different generated and real networks highlight advantages and potential issues for each algorithm.