Research on privacy preserving data mining focus on protecting the sensitive data by transforming the data set and also allow using the data for mining purpose. These techniques help to maintain the accuracy of the results as that of original data set. In the proposed research the data set transformation for protecting the sensitive data is done using the fuzzy member functions. We use various fuzzy member functions such as triangular member function, trapezoidal member function and sigmoid member function to convert the original dataset into perturbed dataset. The accuracy of our research is evaluated by comparing the original data set and perturbed data set on applying to various classification algorithms. The results show that it conserves privacy of sensitive data and also produce valid results.
S. Dhanalakshmi, J. Abdul Samath and M.S. Irfan Ahmed. Model Framework for Sensitive Data Preservation Using Fuzzy.
DOI: https://doi.org/10.36478/ajit.2016.3708.3711
URL: https://www.makhillpublications.co/view-article/1682-3915/ajit.2016.3708.3711