Human Factors in Computing Systems

Meta 📑 Instead of changing the data or learners in multiple ways and then see if fairness improves, postulate that the root causes of bias are the prior decisions that generated the training data. These affect (a) what data was selected, and (b) the labels assigned to the examples. They propose the \(\text{Fair-SMOTE}\) (Fair Synthetic Minority Over Sampling Technique) algorithm which (1) removes biased labels (via situation testing: if the model’s prediction for a data point changes once all of the data points' protected attributes are flipped, then that label is biased and the data point is discarded), and (2) rebalances internal distributions such that based on a protected attribute, examples are equal in both positive and negative classes....

Programmable Money South Korea’s COVID-19 stimulus payments were programmed to expire in Aug 2020, and could only be spent at qualifying shops. Australia’s cashless welfare card [currently in pilot stages] cannot be withdrawn as cash, nor used at alcohol, tobacco and gambling joints. Monzo, a UK-based digital-first mobile baking app, supports IFTTT recipes. This enables budgeting logic, “If I spend at KFC, then move $5 to my penalty pot”....