TY - GEN

T1 - Development of a Personalized Recommendation System with High Data Protection

AU - Palchunova, Olesya

N1 - Conference code: 26

PY - 2025/8/8

Y1 - 2025/8/8

N2 - This paper addresses the development of a personalized recommendation system aimed at improving user experience while ensuring a robust level of data confidentiality. The work describes in detail the critical stages of the recommendation methodology, including the creation of feature vectors representing specific object attributes, the structured collection and secure storage of user interaction datasets, and the client-side computation of personalized preference vectors. A recipe search application implementing the proposed approach is analyzed to demonstrate the efficacy and practical applicability of the developed recommendation algorithms. Particular attention is devoted to the detailed exploration and justification of the recommendation algorithms, data-processing techniques, and privacy preservation mechanisms incorporated within the proposed system. Furthermore, the study discusses the advantages of utilizing client-side computation to mitigate data disclosure risks, thereby facilitating enhanced user trust and compliance with privacy standards. In conclusion, the author outlines potential directions for further research and highlights the possibility for extending and optimizing the presented methodology across other application domains beyond recipe recommendations.

AB - This paper addresses the development of a personalized recommendation system aimed at improving user experience while ensuring a robust level of data confidentiality. The work describes in detail the critical stages of the recommendation methodology, including the creation of feature vectors representing specific object attributes, the structured collection and secure storage of user interaction datasets, and the client-side computation of personalized preference vectors. A recipe search application implementing the proposed approach is analyzed to demonstrate the efficacy and practical applicability of the developed recommendation algorithms. Particular attention is devoted to the detailed exploration and justification of the recommendation algorithms, data-processing techniques, and privacy preservation mechanisms incorporated within the proposed system. Furthermore, the study discusses the advantages of utilizing client-side computation to mitigate data disclosure risks, thereby facilitating enhanced user trust and compliance with privacy standards. In conclusion, the author outlines potential directions for further research and highlights the possibility for extending and optimizing the presented methodology across other application domains beyond recipe recommendations.

KW - logarithmic decay

KW - machine learning

KW - personalization

KW - preference vector

KW - recommendation system

KW - user privacy

UR - https://www.scopus.com/pages/publications/105014202601

UR - https://www.mendeley.com/catalogue/ea40a3c6-cd3e-30b3-a4d7-d0043b792823/

U2 - 10.1109/EDM65517.2025.11096640

DO - 10.1109/EDM65517.2025.11096640

M3 - Conference contribution

SN - 9781665477376

T3 - International Conference of Young Specialists on Micro/Nanotechnologies and Electron Devices, EDM

SP - 1830

EP - 1833

BT - International Conference of Young Specialists on Micro/Nanotechnologies and Electron Devices, EDM

PB - IEEE Computer Society

T2 - 2025 IEEE 26th International Conference of Young Professionals in Electron Devices and Materials (EDM)

Y2 - 27 June 2025 through 1 July 2025

ER -