An Improved Influence Maximization Method for Online Advertising in Social Internet of Things

Area of Science:

• Computer Science
• Network Science
• Artificial Intelligence

Background:

• The Social Internet of Things (SIoT) integrates IoT and social network concepts, finding applications in smart transportation, healthcare, and viral marketing.
• A key challenge in SIoT advertising is identifying optimal diffuser nodes to maximize campaign reach.
• Existing methods struggle to balance influence maximization with cost-efficiency and broad audience engagement.

Purpose of the Study:

• To propose an efficient heuristic algorithm, IMSoT, for influence maximization in SIoT advertising.
• To enhance advertising effectiveness by identifying key nodes while minimizing duplicate advertisements and associated costs.
• To consider the impact of weak nodes for maximizing audience reach in SIoT networks.

Main Methods:

• The IMSoT algorithm employs a two-step approach: candidate object selection and final seed set identification.
• Candidate objects are selected based on degree, local importance, and neighbor analysis (weak and sensitive).
• The final seed set is determined by calculating influence based on overlapping candidate objects to ensure maximum reach and minimal redundancy.

Main Results:

• IMSoT significantly outperforms state-of-the-art algorithms, showing improvements of 38%-193% in considering weak objects.
• The algorithm effectively reduces duplicate advertising costs by 26%-77%.
• IMSoT demonstrates a shorter running time compared to existing algorithms.

Conclusions:

• The IMSoT algorithm provides an efficient and effective solution for influence maximization in SIoT advertising.
• By focusing on weak nodes and preventing ad duplication, IMSoT optimizes reach and reduces costs.
• IMSoT offers a valuable advancement for targeted advertising strategies within the Social Internet of Things ecosystem.