Discrimination Prevention in Data mining with Privacy Preservation

Data mining is an important technology for extracting useful knowledge hidden in large collections of data. There are many positive aspects for data mining, but it also has some disadvantages which include potential discrimination and potential privacy invasion. Discrimination can be defined as treating people unfairingly based on their belonging to a particular group. On the other hand, privacy invasion is the possibility of learning private personal data by unauthorized people. For automated decision making, the classification rules are learned from the training datasets. If the training datasets are biased according to what is discriminatory or sensitive, discriminatory decisions may occur. Discrimination is of two types such as direct discrimination and indirect discrimination. Direct discrimination occurs when decisions are taken based on discriminatory attributes such as religion, race, etc. Indirect discrimination occurs when decisions are made based on attributes that are highly related with the sensitive ones. Anti-discrimination techniques are adopted to avoid or eliminate the discrimination in data mining. While data mining is done, the private data