A multidisciplinary survey on discrimination analysis

Most of the decisions in the today's knowledge society are takenon the basis of historical data by extracting models, patterns,profiles, and rules of human behavior in support of (automated)decision making. There is then the need of developing models,methods and technologies for modelling the processes ofdiscrimination analysis in order to discover and preventdiscrimination phenomena. In this respect, discrimination analysisfrom data should build over the large body of existing legal andeconomic studies. This paper intends to provide amulti-disciplinary survey of the literature on discrimination dataanalysis, including methods for data collection, empiricalstudies, controlled experiments, statistical evidence, and theirlegal requirements and grounds. We cover the following mainstreamresearch lines: labour economic models, (quasi-)experimentalapproaches such as auditing and controlled experiments,profiling-based approaches such as racial profiling and creditmarkets, and the recently blooming research on knowledge discoveryapproaches