Full Length Article
DOI: https://doi.org/10.54216/NIF.040105
Differentially Private Clustering with Dynamic Noise Adjustment (DPC-DNA) based Fusion Anonymity and Privacy Enhancement in Big Data
Other few challenges faced during privacy preservation by anonymity e.g. difficulty in identifying the The main challenges in preserving anonymity for privacy are determining which attributes could undermine privacy and extracting useful information from massive databases without disclosing sensitive details. We developed a Novel Framework for Differentially Private Clustering with Dynamic Noise Adjustment (DPC-DNA) that addresses these issues. This novel approach can recognize sensitive and non-sensitive data aspects using Differentially Private Clustering with Dynamic Noise Adjustment (DPC-DNA). The accuracy of clusters formed by DPC-DNA was assessed using the silhouette score, which gauges how similar each item is to its own group versus others. DPC-DNA achieved a silhouette score of 0.62, signalling strong internal cluster composition. In contrast, traditional k-anonymity clustering yielded a lower score of 0.45, confirming that DPC-DNA significantly boosts accuracy. Our Novel Framework for Differentially Private Clustering with Dynamic Noise Adjustment (DPC-DNA) provides a robust solution for privacy-preserving data mining. By combining differential privacy with adaptive noise management, it safeguards sensitive material while sustaining high precision, integrity and usefulness of results.
Sergey Drominko
visibility
2304
download
3003