Antidiscrimination using Direct and Indirect Methods in Data Mining

Data mining is a very challenging task. It is an important subject in terms of privacy or confidentiality. Discrimination is the act of making a distinction between different things. Discrimination is the act of treating someone differently or unfairly based upon some characteristic. Discrimination can be seen in various places. For example, workplace, school etc. but, everyone has the right to be treated fairly and respectfully. In support of this reason, discrimination removal techniques in data miming have been introduced which includes discrimination discovery and prevention of data. Discrimination deals with direct or indirect discrimination. In direct discrimination sensitive attributes are used, while in indirect discrimination nonsensitive attributes are used for decisions making which are strongly interrelated with biased sensitive data. This approach deals with discrimination prevention and also methodology which is relevant for direct or indirect discrimination prevention individually or together at the same time. Also by using metrics namely MC and GC used to evaluate the effectiveness of the ongoing approach and compare these approaches. Several decision-making tasks are there which let somebody use themselves to become discriminated and helps to preserve good data quality. At the same time direct rule protection methods are combined to achieve better data quality so this combined feature is work efficiently as well as system performance is improved. Improvement is also seen with respect to computational cost of the system, therefore overall system performance gets improved.