A Comparitive Framework For Feature Selction In Privacy Preserving Data Mining Techniques Using Pso And K-Anonumization

The trend of technological era leads to accumulate and utilization of enormous quantity of private details of individuals using internet, which eventually lead to disclose their personal identities. Privacy preserving of data must uphold from revealing sensitive data during the disclosure of the individual's data. Privacy preserving should be incorporated as mining of these datums and the domain deals with this known as Privacy Preserving Data Mining. In the proposed framework, an attribute suppression technique is employed using Particle swarm optimization algorithm and a generalization technique for anonymization is proposed. Also the same work is done using k anonymization and the results are compared for classification accuracy, Precision and recall. In the proposed system Genetic Algorithm and Particle Swarm Optimization takes the common population for evaluation and the results are compared. An optimal generalized feature set is acquired by the PSO and k anonymization technique and is which is used for classification task. The end results of classification are compared with average classification accuracy, average precision and average recall.