Methods for privacy protection using k-anonymity

Large amount of data is produced in electronic form by various governmental and non governmental organizations. This data also has information related to specific individual. Information related to specific individual needs to be protected, so that it may not harm the privacy. Moreover sensitive information related to organization also needs to be protected. Data is released from various organizations as it is demanded by researchers and data mining companies to develop newer and better methods for finding patterns and trends. Any organization who wished to release data has two goals, one is to release the data as close as possible to the original form and second to protect the privacy of individuals and sensitive information from being released. K-anonymity has been used as successful technique in this regard. This method provides a guarantee that released data is at least k-anonymous. Various methods have been suggested to achieve k-anonymity for the given dataset. I categories these methods into four main domains based on the principle these are based and methods they are applying to achieve k-anonymous data. These methods have their respective advantages and disadvantages relating to loss of information, feasibility in real world and suitability to the number of tuples in the dataset.