Generating Private Recommendation System Using Multiple Homomorphic Encryption Scheme

Abstract

The recommender system is important tool in online application to generate the recommendation services. Recommendations are generated by collecting the data from users need; online services access the user’s profiles for generating useful recommendations. Privacy sensitive data is used for to collect the data. Collaborative filtering technique gives privacy for sensitive data if data is misused by other service providers or leaked. Existing system uses Paillier encryption algorithm & DGK algorithm to secure user data from malicious third party as well as to protect the private data against service provider but system is more complex and inefficient. Proposed system protects the privacy of user using encrypting the sensitive data. The system uses multiple homomorphic algorithms to secure user data from service providers. The system is used to protect the confidential data of user against the service provider while providing online services. Encrypting private data is recommended and process on data to generate recommendations. To construct efficient system that does not require the active participation of the user. The experiment shows that the result that provide the security by hiding the personal data of user from third party
DOI: 10.17762/ijritcc2321-8169.150769