Building upon the observation that individuals' decisions to purchase a product are influenced by recommendations from their friends as well as their own preferences, in our work, we propose a new model that factors in people's preferences for a product and the number of his/her neighbors that have adopted this product. In our model, as in related ones, beginning with an active seed set (adopters), an adoption action diffuses in a cascade fashion based on a stochastic rule. We demonstrate that under this model, maximizing individuals' adoption of a product, called the product adoption maximization (PAM) problem, is NP-hard, and the objective function for product adoption is sub-modular for time T (T = 1, 2) when the function for estimating the influence coming from neighbors is sub-linear. Hence, a natural greedy algorithm guarantees an approximation. Furthermore, we show that it is hard to approximate the PAM problem when the function for estimating the influence coming from neighbors is not sub-linear. © 2013 Springer-Verlag Berlin Heidelberg.
CITATION STYLE
Fan, L., Lu, Z., Wu, W., Bi, Y., & Wang, A. (2013). A new model for product adoption over social networks. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7936 LNCS, pp. 737–746). https://doi.org/10.1007/978-3-642-38768-5_67
Mendeley helps you to discover research relevant for your work.