IPrivateSetIntersection
Interfaces · AiDotNet.FederatedLearning.PSI
Defines the interface for Private Set Intersection protocols.
Known implementations: PsiBase
For Beginners
Imagine two hospitals that want to find shared patients without revealing their full patient lists to each other. PSI solves this: each hospital inputs its patient IDs, and the protocol outputs only the IDs that appear in both hospitals’ records, without either hospital learning about patients unique to the other.
How It Works
PSI allows two or more parties to compute the intersection of their sets without revealing elements not in the intersection. This is the foundational building block for entity alignment in vertical federated learning.
All implementations are simulation-safe: they operate on in-memory data structures rather than requiring actual network communication, making them suitable for testing and single-machine VFL experiments.
Properties
| Property | Summary |
|---|---|
string ProtocolName { get; } |
Gets the name of this PSI protocol. |
Methods
| Method | Summary |
|---|---|
PsiResult ComputeIntersection(IReadOnlyList<string> localIds, IReadOnlyList<string> remoteIds, PsiOptions options) |
Computes the intersection between the local party’s IDs and the remote party’s IDs. |
int ComputeCardinality(IReadOnlyList<string> localIds, IReadOnlyList<string> remoteIds, PsiOptions options) |
Computes only the cardinality (count) of the intersection without revealing the actual elements. |