Scholarlink

<p>In the contemporary landscape, the pervasive integration of machine learning and artificial intelligence into our daily routines has resulted in an increased focus on data-centric living. The optimal functionality of these algorithms necessitates the utilization of authentic, voluminous, diverse, and intricate datasets as their primary inputs. However, the reluctance of the general public to furnish such data stems from apprehensions regarding privacy. Thus, there arises a pressing need to devise a robust framework capable of safeguarding data privacy without compromising its utility. Thus, within the framework of our current research, we present an intelligent model meticulously engineered to tackle the intricate privacy challenges inherent in data management. Drawing upon the innate capabilities of artificial intelligence, machine learning, and data correlation analysis, our approach aims to anonymize not only sensitive attributes but also Quasi Identifiers, thereby enhancing privacy protection. Furthermore, our proposed model is intricately tailored to accommodate varying levels of data utility, ensuring that the data remains practically useful. Through a meticulously conducted series of experiments and evaluations, our hybrid model demonstrates superiority over existing algorithms, as evidenced by its outstanding performance, primarily assessed using the Information Gain metric.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Enhancing Privacy Through Cluster-Based Anonymization of Quasi Identifiers in Correlated Datasets

  • 期刊论文

  • 发表日期：

    2025年7月4日

查看全文
  • Sreemoyee Biswas,
  • Kaushiki Tripathi,
  • Nilay Khare,
  • Priyank Jain,
  • Pragati Agrawal,
  • Sanyam Shukla

摘要

In the contemporary landscape, the pervasive integration of machine learning and artificial intelligence into our daily routines has resulted in an increased focus on data-centric living. The optimal functionality of these algorithms necessitates the utilization of authentic, voluminous, diverse, and intricate datasets as their primary inputs. However, the reluctance of the general public to furnish such data stems from apprehensions regarding privacy. Thus, there arises a pressing need to devise a robust framework capable of safeguarding data privacy without compromising its utility. Thus, within the framework of our current research, we present an intelligent model meticulously engineered to tackle the intricate privacy challenges inherent in data management. Drawing upon the innate capabilities of artificial intelligence, machine learning, and data correlation analysis, our approach aims to anonymize not only sensitive attributes but also Quasi Identifiers, thereby enhancing privacy protection. Furthermore, our proposed model is intricately tailored to accommodate varying levels of data utility, ensuring that the data remains practically useful. Through a meticulously conducted series of experiments and evaluations, our hybrid model demonstrates superiority over existing algorithms, as evidenced by its outstanding performance, primarily assessed using the Information Gain metric.