Fairness-aware Data Integration

Abstract

Machine learning can be applied in applications that take decisions that impact people’s lives. Such techniques have the potential to make decision making more objective, but there also is a risk that the decisions can discriminate against certain groups as a result of bias in the underlying data. Reducing bias, or promoting fairness, has been a focus of significant investigation in machine learning, for example based on pre-processing the training data, changing the learning algorithm, or post-processing the results of the learning. However, prior to these activities, data integration discovers and integrates the data that is used for training, and data integration processes have the potential to produce data that leads to biased conclusions. In this paper, we propose an approach that generates schema mappings in ways that take into account: (i) properties that are intrinsic to mapping results that may give rise to bias in analyses; and (ii) bias observed in classifiers trained on the results of different sets of mappings. The approach explores a space of different ways of integrating the data, using a tabu search algorithm, guided by bias-aware objective functions that represent different types of bias.The resulting approach is evaluated using Adult Census and German Credit datasets, to explore the extent to which and the circumstances in which the approach can increase the fairness of the results of the data integration process.

Sociome Data Commons: A scalable and sustainable platform for investigating the full social context and determinants of health

“…These are broadly studied as Social Determinants of Health, which the World Health Organization defines as “conditions in which people are born, grow, live, work and age” and “fundamental drivers of [health] [ 1 ].” Including sociome datasets is often a burdensome data problem, both in finding and integrating disparate datasets, where clinical patient data have to be integrated with other data sources to characterize a patient’s life outside of their clinical interactions. We refer to the entirety of these non-clinical or social factors as a patient’s “sociome.” Due to the diversity of data sources and file types that sociome research has to consider, key bottlenecks in scaling such research to large patient populations include data integration [ 2 ], data harmonization [ 3 ], uneven data quality [ 4 ], and statistical modeling of multimodal datasets [ 5 ]. Consequently, studies often focus on one factor, a composite index, or a set of highly related factors [ 6 ], where potentially crucial nuances and interactions among factors can be lost.…”

“…We refer to the entirety of these non-clinical or social factors as a patient's "sociome." Due to the diversity of data sources and file types that sociome research has to consider, key bottlenecks in scaling such research to large patient populations include data integration [2], data harmonization [3], uneven data quality [4], and statistical modeling of multimodal datasets [5]. Consequently, studies often focus on one factor, a composite index, or a set of highly related factors [6], where potentially crucial nuances and interactions among factors can be lost.…”