This study investigates the effects of K-anonymization on evaluation metrics in Privacy Preserving Data Mining (PPDM) using the Adult dataset. The research proposes the use of the Artificial Bee Colony (ABC) algorithm for feature generalization and suppression, ensuring that features removed do not affect classification accuracy. K-anonymity is achieved through generalization of the original dataset. The Adult dataset, containing 48,842 instances with both categorical and integer attributes from the 1994 Census, is used for evaluation. The study shows that as the k-anonymity level increases, classification accuracy, precision, and recall decrease. The results indicate that the proposed method reduces classification accuracy by up to 1.2308% and recall by up to 1.9209% as anonymity increases from 1 to 90. The study also discusses various PPDM techniques, including randomization, suppression, cryptography, and summarization, and highlights the importance of balancing privacy preservation with data utility. The research contributes to the field of PPDM by providing a framework that effectively addresses privacy concerns while maintaining data utility. The study emphasizes the need for further research to improve the efficiency and effectiveness of PPDM techniques.This study investigates the effects of K-anonymization on evaluation metrics in Privacy Preserving Data Mining (PPDM) using the Adult dataset. The research proposes the use of the Artificial Bee Colony (ABC) algorithm for feature generalization and suppression, ensuring that features removed do not affect classification accuracy. K-anonymity is achieved through generalization of the original dataset. The Adult dataset, containing 48,842 instances with both categorical and integer attributes from the 1994 Census, is used for evaluation. The study shows that as the k-anonymity level increases, classification accuracy, precision, and recall decrease. The results indicate that the proposed method reduces classification accuracy by up to 1.2308% and recall by up to 1.9209% as anonymity increases from 1 to 90. The study also discusses various PPDM techniques, including randomization, suppression, cryptography, and summarization, and highlights the importance of balancing privacy preservation with data utility. The research contributes to the field of PPDM by providing a framework that effectively addresses privacy concerns while maintaining data utility. The study emphasizes the need for further research to improve the efficiency and effectiveness of PPDM techniques.