Inhaltszusammenfassung

We present a novel client/vertical-servers architecture for hybrid multi-party classification problem. The model consists of clients whose attributes are distributed on multiple servers and remain secret during training and testing. Our solution builds privacy-preserving random forests and completes them with a special private set intersection protocol that provides a central commodity server with anonymous conditional statistics. Subsequently, the private set intersection protocol can be use...We present a novel client/vertical-servers architecture for hybrid multi-party classification problem. The model consists of clients whose attributes are distributed on multiple servers and remain secret during training and testing. Our solution builds privacy-preserving random forests and completes them with a special private set intersection protocol that provides a central commodity server with anonymous conditional statistics. Subsequently, the private set intersection protocol can be used to privately classify the queries of new clients using the commodity server’s statistics. The proviso is that the commodity server must not collude with other parties. In cases where this restriction is acceptable, it allows an effective method without computationally expensive public key operations, while it is still secure and avoids precision losses. We report the runtime results on some real-world datasets, and discuss different security aspects and finally give an outlook on further improvements.» weiterlesen» einklappen

Autoren

Klassifikation

DFG Fachgebiet:
4.43 - Informatik

DDC Sachgruppe:
Informatik