Maximizing the Spread of Influence via Generalized Degree Discount

Area of Science:

• Network Science
• Information Diffusion
• Complex Systems

Background:

• Identifying influential spreaders is crucial for controlling information diffusion in networks.
• Existing methods like Degree Discount offer a benchmark but have limitations in real-world network complexity.
• The assumption of treating all nodes equally in Degree Discount is inadequate for diverse network structures.

Purpose of the Study:

• To propose a novel heuristic method, Generalized Degree Discount, as an extension of the Degree Discount method.
• To enhance the identification of influential spreaders by considering a more generalized node status.
• To improve control over spreading processes in real-world networks.

Main Methods:

• Defined node status as the probability of not being influenced by neighbors.
• Introduced a generalized discounted degree index to measure a node's expected influence.
• Sequentially selected spreaders based on their generalized discounted degree in the evolving network.

Main Results:

• Empirical experiments on four real networks demonstrated the effectiveness of the Generalized Degree Discount method.
• Spreaders identified by the proposed approach showed greater influence compared to several benchmark methods.
• The study analyzed the relationship between the new method and common degree-based approaches.

Conclusions:

• The Generalized Degree Discount method is a more effective extension for identifying influential spreaders in complex networks.
• The proposed method offers improved control over spreading processes by accounting for nuanced node characteristics.
• This work advances the understanding of spreader identification in network science.