Differentially Private Gini Index for Income Inequality Analysis

A 693.5 KB dataset from figshare, last updated 2026-05-20, containing results from a study on differentially private computation of the Gini index. The dataset, authored by Wenjie Lan, includes simulated and genuine income data used to evaluate a noise-injection mechanism for protecting confidential microdata. It provides interval estimates for the Gini index and characterizes settings where the method yields accurate results.

Use Cases

• Benchmarking differentially private algorithms based on the described smooth sensitivity mechanism.
• Analyzing the trade-off between privacy and accuracy in inequality metrics based on the described noise injection.
• Developing post-processing methods for private statistics based on the described interval estimation step.
• Studying the local sensitivity of the Gini index based on the described theoretical analysis.

Strengths

• Dataset size is precisely 693.5 KB, indicating a focused scope.
• Includes both simulated and genuine income data as mentioned in the description.
• Mechanism is evaluated against a global sensitivity baseline as described.
• License is clearly defined as CC-BY-4.0.

Limitations

• Row count and column definitions are unknown, limiting suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Data may reflect bias inherent to the specific income data sources used.

Provenance

Source

figshare

Collection Method

Likely contains results from computational experiments and analysis described in the associated paper.

Freshness

Last updated 2026-05-20 15:22:24; freshness should be verified.