In the evolving landscape of machine learning, high-quality data serves as a crucial driver for advancements across industries. With the rise of data transactions, data marketplaces such as AWS Marketplace, Databricks, and Datarade have emerged. However, determining appropriate prices for data products remains a significant challenge due to the unique properties of data products.
Traditional pricing methods can be categorized into cost approach, income approach, and sales comparison approach. The cost approach fails in data pricing due to the near-zero marginal cost from data replication, while the income approach is ineffective due to the inherently unpredictable data revenue. The sales comparison approach remains viable, but its application is hindered by the lack of standardized pricing benchmarks across marketplaces.
To address this challenge, we introduce \texttt{DaDaDa}, the first dataset for data product pricing, containing metadata for 16,147 data products from 9 major data marketplaces worldwide. \texttt{DaDaDa} enables the training of pricing models, thereby establishing price benchmarks for new data products. Additionally, it can be utilized for other important tasks in data markets, such as data product classification and retrieval. Experiments and a retrieval prototype demonstrate the effectiveness of \texttt{DaDaDa} for pricing, classification, and retrieval of data products. The dataset and code are available at GitHub.
Blogger's Review: The introduction of the \texttt{DaDaDa} dataset provides a new perspective on pricing in data markets, addressing the gap of standardized pricing benchmarks. As data transactions continue to grow, this dataset is expected to facilitate more efficient pricing models and smarter data product management.