Eliciting buyer preferences using intelligent agents for multi-attribute dynamic pricing

We address the problem of determining buyer preferences for efficient dynamic pricing by sellers in a competitive online market. Prior reserch on online dynamic pricing by sellers makes certain limiting assumptions such as sellers being aware of buyers preferences, and, prices and profits of competitors, and, buyers selecting products based only on their price. In this paper, we consider a market where buyers and sellers differentiate a product on multiple attributes, and, preferences of different buyers over different product attributes vary temporally. We describe adaptive learning and dynamic pricing algorithms that can be used by a seller to determine buyer preferences and determine a competitive price in the market. These algorithms are encapsulated by software agents that automatically perform the necessary calculations so that the seller can maintain a competitive edge in the market. Simulation results of our algorithms show that they compare favorably against other existing online dynamic pricing algorithms.