Opportunistic work-rest scheduling for productive aging

Abstract

Crowdsourcing platforms are interesting being leveraged by senior citizens for productive aging activities. Algorithmic management (AM) approaches help crowdsourcing systems leverage workers’ intelligence and effort in an optimized manner at scale. However, current AM approaches generally overlook the human aspects of crowdsourcing workers. This prevailing notion has resulted in many existing AM approaches failing to incorporate rest-breaks into the crowdsourcing process to help workers maintain productivity and wellbeing in the long run. To address this problem, we extend the Affective Crowdsourcing (AC) framework to propose the Opportunistic Work-Rest Scheduling (OWRS) approach. It takes into account information on a worker’s mood, current workload and desire to rest to produce dynamic work-rest schedules which jointly minimize collective worker effort output while maximizing collective productivity. Compared to AC, OWRS is able to operate under more diverse mood–productivity mapping functions. As it is a fully distributed approach with time complexity of O(1), it can be implemented as a personal assistant agent for workers. Extensive simulations based on a large-scale real-world dataset demonstrate that OWRS significantly outperforms three baseline scheduling approaches in terms of conserving worker effort while achieving superlinear collective productivity. OWRS establishes a framework which accounts for workers’ heterogeneity to enhance their experience and productivity.