Pricing Strategies for Maximizing Viral Advertising in Social Networks

Viral advertising in social networks is playing an important role for the promotions of new products, ideas and innovations. The advertising usually starts from a set of initial adopters and spreads via social links to become viral. Suppose there is a limited budget, we study optimal pricing strategies to distribute the budget, so that by targeting some initial adopters, the number of people who adopt the advertising can be maximized. The problem of influence maximization have already studied how to select the most influential nodes. However, despite a lot of algorithmic improvement in selecting the most influential users, few of them considered how to incentivize the initial adopters. In this paper, we propose concave probability functions to model the user valuation for sharing the advertising. Given the valuation distributions, we show that finding the optimal pricing strategy is NP-hard. Due to the hardness of this problem, we propose a discrete greedy pricing strategy which has 1 − 1/e − o(1) approximation ratio. We also discuss how to discretize the budget to provide a good trade-off between the performance and the efficiency. Extensive experiments on different data sets are implemented to validate the effectiveness of our algorithm in practice.