Diffusion Mechanism Design in Tree-Structured Social Network
Feiyang Yu
TL;DR
This paper addresses selling multiple items over a tree-structured social network by designing a fixed-price diffusion mechanism that incentivizes buyers to spread information. The mechanism yields individual rationality and diffusion incentive compatibility, and it achieves a worst-case revenue approximation of at least 0.25 compared with the optimal fixed-price auction, with Monte Carlo experiments showing the average performance approaching the optimum as the network grows. The authors provide a branch-based pricing scheme, analyze baseline comparisons, and assess the impact of network structure and a diffusion reward parameter $\alpha$ on revenue. The work offers practical, privacy-preserving revenue gains for sellers and sets the stage for extending diffusion auctions to more general network topologies and scenarios.
Abstract
We design a fixed-price auction mechanism for a seller to sell multiple items in a tree-structured market. The buyers have independently drawn valuation from a uniform distribution, and the seller would like to incentivize buyers to invite more people to the auction. We prove that our mechanism is individual rational, and incentivize compatible with regard to the buyers' action. Furthermore, we show the approximation ratio of our mechanism to the optimal fixed-price auction in two ways, theoretically and via Monte-Carlo simulation, and show a high practical ratio. Finally, we discuss several factors affecting the behavior of our mechanism and its feasibility in reality.