Skip to content

Compare methods

Compare source explanations, strengths, limitations and references for up to four methods.

8 results
Criterion k-Anonymity Differential Privacy Synthetic Data Generation
Overview
  • k-Anonymity is a formal privacy model introduced by Latanya Sweeney in 2002 to protect individuals when personal data is released for research or public use. It requires that every record in a published dataset be indistinguishable from at least k−1 other records with respect to a designated set of quasi-identifying attributes — such as age, gender, and ZIP code — preventing re-identification by linking released data to external sources.
  • Differential privacy is a mathematical framework for releasing statistical information about a dataset while providing rigorous guarantees that individual records cannot be identified or inferred. Introduced by Cynthia Dwork in 2006, it formalizes privacy as a probabilistic bound: any single individual's presence or absence in the dataset changes the output distribution by at most a multiplicative factor of e^ε, where ε is the privacy budget controlling the privacy–utility tradeoff.
  • Synthetic data generation is a statistical disclosure limitation technique introduced by Donald Rubin in 1993, in which values in a confidential dataset are replaced by draws from a fitted posterior predictive distribution rather than released directly. The resulting artificial records preserve the joint statistical structure of the original data while preventing the identification of real individuals, enabling analysts to work with a publicly releasable dataset that behaves like the original for most inferential purposes.
Use and interpretation
  • k-Anonymity is appropriate when tabular microdata must be shared publicly or with external researchers and re-identification through quasi-identifiers is the primary threat. It assumes that the adversary's background knowledge is limited to the quasi-identifier values found in external public datasets. It does not protect against attribute-disclosure attacks (addressed by l-diversity or t-closeness) or adversaries with richer background knowledge. When statistical noise tolerance is preferred over exact suppression, differential privacy may be a stronger alternative.
  • Differential privacy is appropriate when publishing statistics or training machine learning models on sensitive personal data — such as health records, location traces, or financial transactions — where a formal, auditable privacy guarantee is required. It assumes that the analyst can bound query sensitivity and tolerate some accuracy loss from noise. It is not suitable when near-exact individual-level outputs are needed, when ε must be very small and the dataset is tiny, or when non-interactive data release is unnecessary. Alternatives include k-anonymity (weaker guarantee) and secure multi-party computation (when data must stay distributed).
  • Use synthetic data generation when a dataset contains sensitive personal or commercially confidential information that legally or ethically cannot be shared in its original form, yet releasing some version of the data would benefit research or policy analysis. It is well-suited to large, complex microdata files where simple masking techniques would destroy analytical utility. Key assumptions are that the synthesis model captures the relevant joint distribution and that the intended analysis is consistent with what the synthesizer was designed to support. It is a poor fit when the research question depends on individual-level records being authentic, when the synthesis model is badly misspecified, or when adversarial re-identification attacks exploiting subtle distributional artifacts are a serious concern. Alternatives include differential privacy, data enclaves, and query-based access systems.
Output
  • Anonymized table where each record is indistinguishable from at least k-1 others
  • Artificial dataset replacing or augmenting real records
Strengths
  • Provides a mathematically verifiable, deterministic privacy guarantee — every released record is indistinguishable from at least k−1 peers.
  • Preserves the exact structure of the original dataset, making the anonymized table directly usable with standard analytical tools.
  • Transparent and auditable: compliance can be checked by simple record counts per equivalence class without specialized cryptographic tools.
  • Widely adopted in healthcare and census contexts, with mature tooling and regulatory recognition (e.g., HIPAA Safe Harbor guidance).
  • Provides a rigorous, mathematically provable worst-case privacy guarantee independent of attacker background knowledge.
  • Composable: sequential and parallel composition theorems allow privacy budget accounting across complex analysis pipelines.
  • Immune to post-processing: any function applied to a differentially private output remains differentially private.
  • Applicable to both interactive query-answering and non-interactive machine learning (e.g., DP-SGD for neural networks).
  • Releases a fully self-contained dataset that analysts can use with standard software and workflows without special infrastructure.
  • Grounded in a rigorous Bayesian multiple-imputation framework, providing valid variance estimation when combining rules are applied correctly.
  • Can reproduce complex multivariate relationships, rare subgroup distributions, and longitudinal structures that simple masking methods distort.
  • Scales naturally to high-dimensional microdata and can incorporate auxiliary public data to improve the fidelity of the synthesizer.
Limitations
  • Vulnerable to homogeneity attacks: if all k records in an equivalence class share the same sensitive attribute value, the adversary learns that value with certainty.
  • Susceptible to background knowledge attacks when an adversary possesses information beyond the quasi-identifiers used to form equivalence classes.
  • Aggressive generalization can severely reduce data utility, especially for high-dimensional datasets where many attributes interact as quasi-identifiers.
  • Does not quantify the probability of re-identification; the k parameter is a count threshold, not a probabilistic risk measure.
  • Adds noise to query results, reducing statistical accuracy — the tradeoff becomes severe for high-sensitivity queries or small datasets.
  • The privacy budget ε lacks an intuitive absolute scale; choosing a meaningful ε value requires domain expertise and empirical calibration.
  • Naive composition degrades privacy rapidly across many queries; advanced accountants (Rényi DP, moments accountant) are needed for ML workloads.
  • Does not protect against all privacy harms: correlated attributes or auxiliary data sources may still allow inference even under DP.
  • Inferential validity depends critically on correct specification of the synthesis model; a misspecified model produces synthetic data with biased distributions that mislead analysts.
  • Combining rules require knowledge of whether data are fully or partially synthetic and how many replicates were produced, details that must accompany the release.
  • Generating high-fidelity synthetic data for variables with extreme skewness, sparse categories, or complex hierarchical structures remains technically demanding.
  • Membership inference and attribute disclosure attacks have shown that synthetic datasets are not unconditionally private; formal privacy guarantees require additional mechanisms such as differential privacy.
Common pitfalls
  • Selecting too small a k provides weak privacy guarantees while a k that is too large may destroy analytical utility — choosing k requires domain-specific risk assessment.
  • Omitting a quasi-identifier that is available in an external dataset creates a false sense of security; QI identification must be exhaustive and context-aware.
  • Applying k-anonymity to longitudinal or linked datasets without accounting for temporal linkage can allow re-identification across releases.
  • Treating k-anonymity as sufficient for sensitive categories such as health or financial data without layering additional protections (l-diversity, t-closeness, or differential privacy) is inadequate under modern standards.
  • Setting ε too large (e.g., ε > 10) provides negligible practical privacy despite a formal DP label — always interpret ε in context.
  • Ignoring sensitivity unboundedness: queries on unbounded domains (raw sums without clipping) yield infinite sensitivity and break the mechanism.
  • Treating DP as a complete privacy solution: it bounds individual disclosure risk but does not prevent group-level or aggregate inferences.
  • Failing to account for composition across the entire data pipeline; each query or model training step consumes budget and must be tracked.
  • Releasing only a single synthetic replicate (M = 1) invalidates the combining-rules framework and produces overconfident standard errors.
  • Fitting the synthesis model on a non-representative subsample and then generating synthetic records at population scale introduces systematic bias that analysts may not detect.
  • Treating synthetic data as a perfect substitute for real data in analyses that require individual authenticity — such as record linkage or geographic micro-targeting — leads to substantively wrong conclusions.
  • Omitting rare but analytically important subgroups from the synthesis model can cause those groups to be absent or severely distorted in the synthetic output.
Introduced by
  • Latanya Sweeney
  • Cynthia Dwork
  • Donald Rubin
Year
  • 2002
  • 2006
  • 1993