KNN classification service system and method supporting privacy protection

Abstract

The invention belongs to the field of machine learning and privacy protection, and particularly relates to a KNN classification service system and method supporting privacy protection. The architecture of the system comprises a model owner and a client; the method of the KNN classification service system supporting privacy protection comprises the following steps: 1) a preparation stage: generating a public key and a private key, and encrypting training data according to the public key; 2) a classification stage: two parties interact with keys; and the client encrypts to-be-tested data throughthe public key, the model owner completes encrypted data classification by cooperating with the client through a security protocol based on the encrypted training data, and finally obtains a classification result and sends the classification result to the client. According to the method, training data and to-be-tested data are encrypted by using homomorphic encryption calculation, a secure basicprotocol is constructed by combining a secure multi-party calculation technology and homomorphic encryption, and a secure KNN classifier is constructed based on the secure basic protocol, so that thetwo parties realize analysis and prediction of personal data on the premise of ensuring that the privacy of the personal data is not leaked.

Description

technical field [0001] The invention belongs to the field of machine learning and privacy protection, and in particular relates to a KNN classification service system and method supporting privacy protection. Background technique [0002] The KNN classification service, that is, the k-Nearest Neighbor (KNN) classifier may leak user privacy information during the sample training and classification stages. In the sample training stage, the data owner does not want the data information he owns to be leaked out, and even keep the trainer confidential, which requires encryption of the training data. In the classification stage, the trainer will use the obtained model W as a component of the classifier, and publish the classifier to provide services, but does not want the results to be obtained by a third party, which requires encryption of the classification model and test vector. Therefore, for classifiers, whether it is the training phase or the classification phase, the issue...

AI Technical Summary

Problems solved by technology

[0006] Although the combination of homomorphic encryption scheme and secure multi-party computation can partially solve the problem of classifier data privacy protection, and there have been some research results on classifier privacy protection, but there are still the following problems: 1) Most of the schemes are aimed at training There is little protection for the classification model and classification process for the privacy protection of stage data; 2) Its security settings are low, and it is easy to leak the classification model and classification results; 3) Homomorphic encryption operations support polynomial operations of addition and multiplication , the comparison operation can also be obtained through secure multi-party computing, but the machine learning calculation overhead is large and the efficiency is low

Images