Gradient-Guided Credit Assignment and Joint Optimization for Dependency-Aware Spatial Crowdsourcing
Yafei Li, Wei Chen, Jinxing Yan, Huiling Li, Lei Gao, Mingliang Xu
Abstract
Dependency-aware spatial crowdsourcing (DASC) addresses the unique challenges posed by subtask dependencies in spatial task assignment. This paper investigates the task assignment problem in DASC and proposes a two-stage Recommend and Match Optimization (RMO) framework, leveraging multi-agent reinforcement learning for subtask recommendation and a multi-dimensional utility function for subtask matching. The RMO framework primarily addresses two key challenges: credit assignment for subtasks with interdependencies and maintaining overall coherence between subtask recommendation and matching. Specifically, we employ meta-gradients to construct auxiliary policies and establish a gradient connection between two stages, which can effectively address credit assignment and joint optimization of subtask recommendation and matching, while concurrently accelerating network training. We further establish a unified gradient descent process through gradient synchronization across recommendation networks, auxiliary policies, and the matching utility evaluation function. Experiments on two real-world datasets validate the effectiveness and feasibility of our proposed approach.
BibTeX
@article{Li_Chen_Yan_Li_Gao_Xu_2025, title={Gradient-Guided Credit Assignment and Joint Optimization for Dependency-Aware Spatial Crowdsourcing}, volume={39}, url={https://ojs.aaai.org/index.php/AAAI/article/view/33566}, DOI={10.1609/aaai.v39i13.33566}, abstractNote={Dependency-aware spatial crowdsourcing (DASC) addresses the unique challenges posed by subtask dependencies in spatial task assignment. This paper investigates the task assignment problem in DASC and proposes a two-stage Recommend and Match Optimization (RMO) framework, leveraging multi-agent reinforcement learning for subtask recommendation and a multi-dimensional utility function for subtask matching. The RMO framework primarily addresses two key challenges: credit assignment for subtasks with interdependencies and maintaining overall coherence between subtask recommendation and matching. Specifically, we employ meta-gradients to construct auxiliary policies and establish a gradient connection between two stages, which can effectively address credit assignment and joint optimization of subtask recommendation and matching, while concurrently accelerating network training. We further establish a unified gradient descent process through gradient synchronization across recommendation networks, auxiliary policies, and the matching utility evaluation function. Experiments on two real-world datasets validate the effectiveness and feasibility of our proposed approach.}, number={13}, journal={Proceedings of the AAAI Conference on Artificial Intelligence}, author={Li, Yafei and Chen, Wei and Yan, Jinxing and Li, Huiling and Gao, Lei and Xu, Mingliang}, year={2025}, month={Apr.}, pages={14301-14308} }