Toward a secure metaverse: Crafting cutting-edge algorithm for protected data analysis

To propel the metaverse into a more advanced realm, it is imperative to leverage realworld data analysis is imperative. Databases containing personal information harbor vast potential and find widespread application in academic and corporate settings. Extensive research on privacy protection that recognizes the paramount importance of privacy considerations when accessing such databases, is ongoing. Traditionally, anonymization serves as the foundation for privacy protection by excluding identifiers capable of uniquely pinpointing individuals and adjusting quasi-identifiers (QIDs). However, prevailing techniques predominantly focus on modifying QIDs, sometimes resulting in substantial alterations that make them unsuitable for analysis depending on the database or technology. In this chapter, we introduce an anonymization algorithm focused on augmenting dummy records without altering QIDs. Our study's innovation lies in minimizing the number of dummy records introduced during anonymization while maintaining safety through l-diversity, a pivotal privacy metric, and demonstrating superior utility compared to existing methods. Through database-driven experiments, we elucidate a notable trade-off between safety and utility compared to existing techniques. By facilitating the ethical utilization of personal information, this technology stands poised to elevate the metaverse to unprecedented sophistication.