PMA for Privacy Preservation in Data Mining

Privacy is becoming a progressively important issue in many data mining applications. This has initiated the development of many privacy preserving data mining techniques. In recent years, various data mining algorithms combining privacy preserving techniques have been established that hide sensitive identifiers or patterns. When applying privacy preservation techniques, importance is given to the utility and information loss. In this paper we propose Statistical Disclosure Control (SDC) based Perturbed Micro Aggregation (PMA) for anonymizing the individual records. Through the experimental results, the proposed technique is validated to prevent the disclosure of sensitive data without degradation of data utilization. Our work highlights some discussions about future work and promising directions in the perspective of privacy preservation in data mining. Keywords—PPDM; privacy; microaggregation; microdata; anonymization; data mining