Equilibrium of Data Markets with Externality

Authors: Safwan Hossain, Yiling Chen

ICML 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We experimentally validate our theoretical insights on the effectiveness of transaction costs. While no publicly available dataset exists for data markets with prices and utilities, we take inspiration from AWS marketplace (a fixed price data market platform) we design a suite of synthetic experiments.
Researcher Affiliation Academia 1Harvard University, Cambridge, USA. Correspondence to: Safwan Hossain <shossain@g.harvard.edu>.
Pseudocode Yes Algorithm 1 Zooming algorithm for buyer i
Open Source Code No The paper mentions designing 'a suite of synthetic experiments' and uses 'a dataset derived from AWS Data Exchange' but does not provide any links to source code or explicit statements about its release.
Open Datasets No On a dataset derived from AWS Data Exchange. This is a derived dataset for synthetic experiments, and no public access details (link, DOI, specific citation to a public version) are provided for this specific dataset.
Dataset Splits No The paper does not provide specific training, validation, or test split percentages or sample counts for any dataset.
Hardware Specification No The paper does not provide specific details about the hardware used for running the experiments.
Software Dependencies No The paper does not provide specific software dependencies with version numbers.
Experiment Setup Yes We take the 10 different data categories from AWS, and instantiate several data sellers for each category (177 total sellers). For each category, we have several buyers (57 total), where each buyer has zero gain/externality for the sellers not in their category, with gains and externalities for their category sampled uniformly. They may purchase from up to 10% of the sellers in their category.