| Summary: | 碩士 === 國立彰化師範大學 === 資訊管理學系所 === 98 === Consumers often form complex social networks based on a multitude of different relations and interactions. By virtue of these interactions, they influence each other’s decisions in adopting a product or behavior. Viral marketing takes advantage of these social network effects, wherein friends recommend a product to their friends, who in turn recommend it to others, and so forth, creating a cascade of recommendation. Given a social network, a natural question thus emerges in the area of viral marketing: If we wish to maximize the size of a cascade, which subset of consumers should we target? This influence maximization problem has been proposed and studied in previous researches. Though previous researches shed light to this interesting problem, their views of social networks are highly idealistic. First, previous researches assume that a social network is static. Hence it is an interesting question to consider how a company may effectively infiltrate a market in which the structure of consumer network is evolving and a stronger competing company is present. In this research, we utilize the search and adaption capacity of ant colony optimization (ACO) algorithm to solve the influence maximization problem. The proposed approaches are evaluated by using real-world data and the experimental results show that Node’s Influence approach has the best performance in competitive and dynamic social networks.
|
|---|
