An Empirical Study on Preserving Sensitive Knowledge in Data Mining

Protection of data in data mining involves preserving sensitive data and sensitive knowledge. Preserving sensitive data focus on protecting the data when it is shared for mining with third parties. In some situations even the result of data mining techniques might reveal sensitive information. This area of research which focuses on output preservation of data mining techniques is termed as Preserving Sensitive Knowledge. The sensitive results of data mining are preserved by reducing the efficiency of the result. In this paper, an empirical study on existing sensitive knowledge preservation approaches is discussed. The approaches include rule hiding in association rule mining and classification technique. The output data protection also includes the restriction in query processing. The article highlights the merits and demerits on each approach.