IDENTIFICATION AND MITIGATION OF ALGORITHMIC BIAS THROUGH POLICY INSTRUMENTS

With the increasing implementation of technologies like Artificial Intelligence, Machine Learning and sophisticated data analysis algorithms, the process of decision-making and recommendations is being automated with more emphasis on a trained system to generate valuable required outcomes. Taking into consideration the aspect of how the mechanism of any computer-based decision-making process takes the order, this paper aims at understanding the possibility of bias in any output, which may, further, become a cause for bias in the real world. This paper tries to address a stepwise approach to discovering the factors of how such a bias can be quantified and presents a view of the impact parameters for various stakeholders while trying to process an approach for the overall fairness of the system.