IDP - OPTICS: Improvement of Differential Privacy Algorithm in Data Histogram Publishing Based on Density Clustering

In recent years, incidents of privacy leaks have frequently occurred. How to protect the group and personal privacy has become a focus issue in the field of information security. Differential privacy technology can protect the privacy of published data by adding noise. In this paper, we propose the IDP-OPTICS algorithm to solve the problem of excessive noise accumulation and low data availability in the histogram data release under the static data set. First, we preprocess the data with a density-based clustering algorithm. Then, we apply differential privacy to the data and finally add noise to the processed data by adding heteroscedasticity. The experimental data shows that IDP-OPTICS achieves a reduction in the problem of excessive noise accumulation and a balance of noise distribution as a whole.