Discrimination Discovery and Prevention in Data Mining

Data mining is an important technology for extracting useful knowledge in large collections of data. However, the perceptions about data mining are negative as potential privacy invasion and potential discrimination. Unjustified distinction of individuals based on their membership in a certain group or category are referred as discrimination. At first sight, automated data collection and data mining techniques such as classification rule mining may give a sense of fairness and may not guide themselves by personal preferences. However, classification rules are actually learned by the system from the training data and training data sets itself are biased in what regards discriminatory (sensitive) attributes like gender, race, religion, etc. To extract knowledge without violation such as privacy and non-discrimination is most difficult and challenging. The discrimination discovery and prevention techniques have been discussed in this papers.