PS-MI: Accurate, Efficient, and Private Data Valuation in Vertical Federated Learning

Summary: PS-MI performs model-agnostic VFL data valuation by estimating coalition–label mutual information with private random projections. Stratified coalition sampling, LSH, batching, and early stopping make kNN estimation accurate and up to 592× faster than prior VFDV methods. (summarized by gpt-5.6-luna on Jul 24 2026)

Paper ID: ha1ab45c19f37d5f2
Venue: VLDB
Year: 2025
Pagerank: 4.9605343e-05
Overall Rank: 11,386 | 23.68%
DOI: 10.14778/3748191.3748215
PDF: Download (CC BY-NC-ND 4.0)

Incoming Non-self Citations Over Time

No non-self incoming citations found for this paper in this database.

Authors

1. Xiaokai Zhou (Wuhan University)
2. Xiao Yan (Wuhan University)
3. Fangcheng Fu (Shanghai Jiao Tong University)
4. Ziwen Fu (Wuhan University)
5. Tieyun Qian (Wuhan University)
6. Yuanyuan Zhu (Wuhan University)
7. Qinbo Zhang (Wuhan University)
8. Bin Cui (Peking University)
9. Jiawei Jiang (Wuhan University)

BibTeX Citation

@article{zhou_vldb25,
        title = {{PS-MI: Accurate, Efficient, and Private Data Valuation in Vertical Federated Learning}},
        author = {Zhou, Xiaokai and Yan, Xiao and Fu, Fangcheng and Fu, Ziwen and Qian, Tieyun and Zhu, Yuanyuan and Zhang, Qinbo and Cui, Bin and Jiang, Jiawei},
        journal = {PVLDB},
        series = {{VLDB} '25},
        volume = {18},
        number = {10},
        pages = {3559--3572},
        doi = {10.14778/3748191.3748215},
        url = {https://doi.org/10.14778/3748191.3748215},
        year = {2025}
}

Incoming Citations (Sorted by Pagerank)

Showing 0 of 0 citing papers.

Rank	Citing Paper	Year	Venue	Pagerank

Outgoing Citations (Sorted by Pagerank)

Showing 16 of 16 cited papers.

Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.

Rank	Cited Paper	Year	Venue	Pagerank
170	Combining Fuzzy Information from Multiple Systems	1996	PODS	0.0002707332
897	Efficient Task-Specific Data Valuation for Nearest Neighbor Algorithms	2019	VLDB	0.00013162662
1,284	Privacy Preserving Vertical Federated Learning for Tree-based Models	2020	VLDB	0.00011195435
2,020	VF2Boost: Very Fast Vertical Federated Gradient Boosting for Cross-Enterprise Learning	2021	SIGMOD	9.1641863e-05
3,324	BlindFL: Vertical Federated Machine Learning without Peeking into Your Data	2022	SIGMOD	7.4111459e-05
3,694	Secure Shapley Value for Cross-Silo Federated Learning	2023	VLDB	7.0721664e-05
3,947	Understanding and Benchmarking the Impact of GDPR on Database Systems	2020	VLDB	6.8897508e-05
4,955	Efficient Sampling Approaches to Shapley Value Approximation	2023	SIGMOD	6.3262068e-05
5,736	Federated Matrix Factorization with Privacy Guarantee	2022	VLDB	5.995594e-05
5,789	OpBoost: A Vertical Federated Tree Boosting Framework Based on Order-Preserving Desensitization	2023	VLDB	5.9762627e-05
6,486	Differentially Private Vertical Federated Clustering	2023	VLDB	5.7522634e-05
6,508	Practical Differentially Private and Byzantine-resilient Federated Learning	2023	SIGMOD	5.7451686e-05
6,566	Falcon: A Privacy-Preserving and Interpretable Vertical Federated Learning System	2023	VLDB	5.728236e-05
8,233	Contributions Estimation in Federated Learning: A Comprehensive Experimental Evaluation	2024	VLDB	5.3571186e-05
9,689	FEAST: A Communication-efficient Federated Feature Selection Framework for Relational Data	2023	SIGMOD	5.1258854e-05
10,376	Towards Communication-efficient Vertical Federated Learning Training via Cache-enabled Local Updates	2022	VLDB	4.9979062e-05

Semantically Similar Papers

#	Overall Rank	Paper	Year	Venue
1	3,694	Secure Shapley Value for Cross-Silo Federated Learning	2023	VLDB
2	5,882	Federated Heavy Hitter Analytics with Local Differential Privacy	2025	SIGMOD
3	2,020	VF2Boost: Very Fast Vertical Federated Gradient Boosting for Cross-Enterprise Learning	2021	SIGMOD
4	3,324	BlindFL: Vertical Federated Machine Learning without Peeking into Your Data	2022	SIGMOD
5	1,284	Privacy Preserving Vertical Federated Learning for Tree-based Models	2020	VLDB
6	8,233	Contributions Estimation in Federated Learning: A Comprehensive Experimental Evaluation	2024	VLDB
7	5,736	Federated Matrix Factorization with Privacy Guarantee	2022	VLDB
8	3,319	FedKNN: Secure Federated k-Nearest Neighbor Search	2024	SIGMOD
9	4,558	Projected Federated Averaging with Heterogeneous Differential Privacy	2022	VLDB
10	6,486	Differentially Private Vertical Federated Clustering	2023	VLDB