Equity Promotion in Online Resource Allocation

We consider online resource allocation under a typical non-profit setting, where limited or even scarce resources are administered by not-for-profit organization like government. focus on the internal-equity assuming that arriving requesters homogeneous in terms of their external factors demands but heterogeneous for internal attributes demographics. Specifically, we associate each requester with one several groups based demographics (i.e., race, gender, and age), aim to design an equitable distributing strategy such every group can receive fair share proportional preset target ratio. present two LP-based sampling algorithms investigate them both theoretically (in competitive-ratio analysis) experimentally real COVID-19 vaccination data maintained Minnesota Department Health. Both theoretical numerical results show our strategies effectively promote equity, especially when arrival population is disproportionately represented, as observed early stage vaccine rollout.