mlsanityData to 7 Oct 2026

Data Poisoning Attacks and Defenses to Crowdsourcing Systems

arXiv 2102.09171 ↗First posted 18 Feb 2021cs.CR cs.DC cs.HCcited by 12

On the map: LLM safety and security → Backdoor and data poisoning attacks

Abstract (arXiv)

A key challenge of big data analytics is how to collect a large volume of (labeled) data. Crowdsourcing aims to address this challenge via aggregating and estimating high-quality data (e.g., sentiment label for text) from pervasive clients/users. Existing studies on crowdsourcing focus on designing new methods to improve the aggregated data quality from unreliable/noisy clients. However, the security aspects of such crowdsourcing systems remain under-explored to date. We aim to bridge this gap in this work. Specifically, we show that crowdsourcing is vulnerable to data poisoning attacks, in which malicious clients provide carefully crafted data to corrupt the aggregated data. We formulate our proposed data poisoning attacks as an optimization problem that maximizes the error of the aggregated data. Our evaluation results on one synthetic and two real-world benchmark datasets demonstrate that the proposed attacks can substantially increase the estimation errors of the aggregated data. We also propose two defenses to reduce the impact of malicious clients. Our empirical results show that the proposed defenses can substantially reduce the estimation errors of the data poisoning attacks.

Key facts. Cited by 12 papers in the map (data through 7 Oct 2026). Placed in Backdoor and data poisoning attacks, one of 12 directions in LLM safety and security. Builds on 14 of its 49 references that are in the map. First posted on arXiv 18 Feb 2021.

Builds on

14 of 49 references.

References · 14

  1. 2017Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoningcited by 1,012
  2. 2018Poison Frogs! Targeted Clean-Label Poisoning Attacks on Neural Networkscited by 500
  3. 2019Local Model Poisoning Attacks to Byzantine-Robust Federated Learningcited by 479
  4. 2018Analyzing Federated Learning through an Adversarial Lenscited by 346
  5. 2020FLTrust: Byzantine-robust Federated Learning via Trust Bootstrappingcited by 330
  6. 2019A Little Is Enough: Circumventing Defenses For Distributed Learningcited by 285
  7. 2018Manipulating Machine Learning: Poisoning Attacks and Countermeasures for Regression Learningcited by 219
  8. 2019Fall of Empires: Breaking Byzantine-tolerant SGD by Inner Product Manipulationcited by 136
  9. 2018Is feature selection secure against training data poisoning?cited by 88
  10. 2016Data Poisoning Attacks on Factorization-Based Collaborative Filteringcited by 78
  11. 2020Influence Function based Data Poisoning Attacks to Top-N Recommender Systemscited by 61
  12. 2018Poisoning Attacks to Graph-Based Recommender Systemscited by 52
2 more
  1. 2015A Survey on Truth Discoverycited by 36
  2. 2015Truth Finding on the Deep Web: Is the Problem Solved?cited by 10

Built on by

Cited by year

Citing papers per year: 2021: 1; 2022: 3; 2023: 1; 2024: 4; 2025: 3.

12 papers.

  1. 2024Robustifying Safety-Aligned Large Language Models through Clean Data Curationcited by 43
  2. 2025Practical Poisoning Attacks against Retrieval-Augmented Generationcited by 35
  3. 2022PoisonedEncoder: Poisoning the Unlabeled Pre-training Data in Contrastive Learningcited by 33
  4. 2022AFLGuard: Byzantine-robust Asynchronous Federated Learningcited by 29
  5. 2024Poisoning Federated Recommender Systems with Fake Userscited by 15
  6. 2024Tracing Back the Malicious Clients in Poisoning Attacks to Federated Learningcited by 6
  7. 2022FairRoad: Achieving Fairness for Recommender Systems with Optimized Antidote Datacited by 5
  8. 2023Competitive Advantage Attacks to Decentralized Federated Learningcited by 2
  9. 2021Classification Auto-Encoder based Detector against Diverse Data Poisoning Attackscited by 1
  10. 2024Robust Decision Aggregation with Adversarial Expertscited by 1
  11. 2025FL-PLAS: Federated Learning with Partial Layer Aggregation for Backdoor Defense Against High-Ratio Malicious Clientscited by 1
  12. 2025Clean Image May be Dangerous: Data Poisoning Attacks Against Deep Hashingcited by 0