GKD-Recruiter: Jointly Modeling Social and Task Heterogeneity for Spatial Crowdsourcing via Graph Knowledge Distillation

Yucen Gao, Zhemeng Yu, Zhuoran Li, Jianxiong Guo, Xiaofeng Gao
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:33826-33840, 2026.

Abstract

Social recruitment offers a solution to worker scarcity in Spatial Crowdsourcing (SC) but faces challenges that are often ignored in traditional Influence Maximization. First, task heterogeneity arising from offline execution constraints breaks the “interest-implies-participation” assumption, as social influence often fails to translate into physical presence. Second, finite task demand creates a “saturation trap”, a non-submodular setting in which utility drops sharply to zero once demand is met. To bridge these gaps, we propose GKD-Recruiter, a Task-Aware framework designed to maximize Effective Task Satisfaction (ETS). We explicitly model the complex worker-task affinity via a heterogeneous graph and capture directional social influence using a novel Influential GAT. To robustly fuse these distinct signals, we introduce a Graph Knowledge Distillation mechanism. Furthermore, we employ Rainbow DQN to navigate the non-submodular combinatorial search space, avoiding the local optima that trap greedy heuristics. Extensive experiments on the real-world dataset demonstrate that GKD-Recruiter significantly outperforms state-of-the-art baselines in both solution quality and inference efficiency. The code is available at https://github.com/GaoYucen/GKD-Recruiter.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-gao26ai, title = {{GKD}-Recruiter: Jointly Modeling Social and Task Heterogeneity for Spatial Crowdsourcing via Graph Knowledge Distillation}, author = {Gao, Yucen and Yu, Zhemeng and Li, Zhuoran and Guo, Jianxiong and Gao, Xiaofeng}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {33826--33840}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/gao26ai/gao26ai.pdf}, url = {https://proceedings.mlr.press/v306/gao26ai.html}, abstract = {Social recruitment offers a solution to worker scarcity in Spatial Crowdsourcing (SC) but faces challenges that are often ignored in traditional Influence Maximization. First, task heterogeneity arising from offline execution constraints breaks the “interest-implies-participation” assumption, as social influence often fails to translate into physical presence. Second, finite task demand creates a “saturation trap”, a non-submodular setting in which utility drops sharply to zero once demand is met. To bridge these gaps, we propose GKD-Recruiter, a Task-Aware framework designed to maximize Effective Task Satisfaction (ETS). We explicitly model the complex worker-task affinity via a heterogeneous graph and capture directional social influence using a novel Influential GAT. To robustly fuse these distinct signals, we introduce a Graph Knowledge Distillation mechanism. Furthermore, we employ Rainbow DQN to navigate the non-submodular combinatorial search space, avoiding the local optima that trap greedy heuristics. Extensive experiments on the real-world dataset demonstrate that GKD-Recruiter significantly outperforms state-of-the-art baselines in both solution quality and inference efficiency. The code is available at https://github.com/GaoYucen/GKD-Recruiter.} }
Endnote
%0 Conference Paper %T GKD-Recruiter: Jointly Modeling Social and Task Heterogeneity for Spatial Crowdsourcing via Graph Knowledge Distillation %A Yucen Gao %A Zhemeng Yu %A Zhuoran Li %A Jianxiong Guo %A Xiaofeng Gao %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-gao26ai %I PMLR %P 33826--33840 %U https://proceedings.mlr.press/v306/gao26ai.html %V 306 %X Social recruitment offers a solution to worker scarcity in Spatial Crowdsourcing (SC) but faces challenges that are often ignored in traditional Influence Maximization. First, task heterogeneity arising from offline execution constraints breaks the “interest-implies-participation” assumption, as social influence often fails to translate into physical presence. Second, finite task demand creates a “saturation trap”, a non-submodular setting in which utility drops sharply to zero once demand is met. To bridge these gaps, we propose GKD-Recruiter, a Task-Aware framework designed to maximize Effective Task Satisfaction (ETS). We explicitly model the complex worker-task affinity via a heterogeneous graph and capture directional social influence using a novel Influential GAT. To robustly fuse these distinct signals, we introduce a Graph Knowledge Distillation mechanism. Furthermore, we employ Rainbow DQN to navigate the non-submodular combinatorial search space, avoiding the local optima that trap greedy heuristics. Extensive experiments on the real-world dataset demonstrate that GKD-Recruiter significantly outperforms state-of-the-art baselines in both solution quality and inference efficiency. The code is available at https://github.com/GaoYucen/GKD-Recruiter.
APA
Gao, Y., Yu, Z., Li, Z., Guo, J. & Gao, X.. (2026). GKD-Recruiter: Jointly Modeling Social and Task Heterogeneity for Spatial Crowdsourcing via Graph Knowledge Distillation. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:33826-33840 Available from https://proceedings.mlr.press/v306/gao26ai.html.

Related Material