Fair knapsack pricing for data marketplaces

Abstract

Data has become an important economic good. This has led to the development of data marketplaces which facilitate trading by bringing data vendors and data consumers together on one platform. Despite the existence of such infrastructures, data vendors struggle to determine the value their offerings have to customers. This paper explores a novel pricing scheme that allows for price discrimination of customers by selling custom-tailored variants of a data product at a price suggested by a customer. To this end, data quality is adjusted to meet a customer’s willingness to pay. To balance customer preferences and vendor interest, a model is developed, translating fair pricing into a Multiple-Choice Knapsack Problem and making it amenable to an algorithmic solution.