Data Measurements for Decentralized Data Markets

TL;DR

This paper introduces federated data measurement techniques for decentralized data markets, enabling buyers to efficiently identify relevant and diverse datasets from sellers without intermediaries or specialized models.

Contribution

It proposes and benchmarks novel federated data measurement methods that facilitate seller selection based on diversity and relevance in decentralized markets.

Findings

Federated measurements effectively identify relevant datasets

Diversity measures improve dataset selection quality

Benchmark results demonstrate practical viability

Abstract

Decentralized data markets can provide more equitable forms of data acquisition for machine learning. However, to realize practical marketplaces, efficient techniques for seller selection need to be developed. We propose and benchmark federated data measurements to allow a data buyer to find sellers with relevant and diverse datasets. Diversity and relevance measures enable a buyer to make relative comparisons between sellers without requiring intermediate brokers and training task-dependent models.