On the task assignment with group fairness for spatial crowdsourcing

Task assignment, the core problem of Spatial Crowdsourcing (SC), is often modeled as an optimization with multiple constraints, and quality efficiency its solution determines how well SC system works. Fairness a critical aspect task assignment that affects worker participation satisfaction. Although existing studies on have noticed fairness problem, they mainly focus at individual level rather than group level. However, differences among groups in certain attributes (e.g. race, gender, age) can easily lead to discrimination which triggers ethical issues even deteriorates system. Therefore, we study for SC. According principle fair budget allocation, define well-designed constraint be considered systems, resulting fairness. We consider common One-to-One (O2-SC), our goal maximize while satisfying other constraints such spatial constraints. Specifically, first give formal definition O2-SC. Then, prove it essentially NP-hard combinatorial problem. Next, provide novel fast algorithm theoretical guarantees solve it. Finally, conduct extensive experiments using both synthetic real datasets. The experimental results show proposed significantly improve algorithms, completely random algorithm. also efficiently effectively complete systems ensuring