Measure context-dependent economic value of data

Determine measurement and valuation frameworks for datasets whose economic worth depends on buyer-specific context, existing data holdings, downstream application, composition with other datasets, and access exclusivity, and develop price-discovery mechanisms that account for interdependent preferences and the inspection–copying paradox in data exchange.

Background

The paper argues that data’s value is highly context-dependent and compositional: the same dataset can have different value depending on a buyer’s existing corpora, intended use, and whether access is exclusive or widespread. These features undermine uniform pricing and complicate market formation due to adverse selection and the verification paradox (inspection enables copying).

Section 6 elaborates that technical units (tokens, records) do not map cleanly to economic value and that quality is task-specific. It calls for measurement approaches that capture context dependence and price-discovery mechanisms robust to copying risks.