Paper 2023/442

Non-interactive privacy-preserving naive Bayes classifier using homomorphic encryption

Jingwei Chen, Chongqing Institute of Green and Intelligent Technology, CAS
Yong Feng, Chongqing Institute of Green and Intelligent Technology, CAS
Yang Liu, Chongqing Jiaotong University
Wenyuan Wu, Chongqing Institute of Green and Intelligent Technology, CAS
Guanci Yang, Guizhou University

In this paper, we propose a non-interactive privacy-preserving naive Bayes classifier from leveled fully homomorphic encryption schemes. The classifier runs on a server that is also the model’s owner (modeler), whose input is the encrypted data from a client. The classifier produces encrypted classification results, which can only be decrypted by the client, while the modelers model is only accessible to the server. Therefore, the classifier does not leak any privacy on either the servers model or the clients data and results. More importantly, the classifier does not require any interactions between the server and the client during the classification phase. The main technical ingredient is an algorithm that computes the maximum index of an encrypted array homomorphically without any interactions. The proposed classifier is implemented using HElib. Experiments show the accuracy and efficiency of our classifier. For instance, the average cost can achieve about 34ms per sample for a real data set in UCI Machine Learning Repository with the security parameter about 100 and accuracy about 97%.

Available format(s)
Publication info
Published elsewhere. the 4th EAI International Conference on Security and Privacy in New Computing Environments
privacy-preserving machine learningnaïve Bayes classifierfully homomorphic encryptionBGVHElib
Contact author(s)
jingwei chen @ outlook com
2023-03-27: approved
2023-03-27: received
See all versions
Short URL
Creative Commons Attribution


      author = {Jingwei Chen and Yong Feng and Yang Liu and Wenyuan Wu and Guanci Yang},
      title = {Non-interactive privacy-preserving naive Bayes classifier using homomorphic encryption},
      howpublished = {Cryptology ePrint Archive, Paper 2023/442},
      year = {2023},
      doi = {10.1007/978-3-030-96791-8_14},
      note = {\url{}},
      url = {}
Note: In order to protect the privacy of readers, does not use cookies or embedded third party content.