Abstract:Crowdsourcing provides open platforms where people worldwide can post or participate in some tasks. However, its open feature also creates security and privacy issues. In this paper, we first demonstrate that malicious attackers can infer some participants' sensitive information by linking their responses to multiple seemingly unrelated crowdsourcing survey tasks. What makes the linkage attack in crowdsourcing different from the traditional privacy-preserved data publishing is that the data collectors are malicious: they can immediately see a survey participant's responded information once submitted and even can intentionally design "follow-up" surveys to dig more data about the target user. To protect against this new vulnerability in crowdsourcing, we propose a personal Privacy Monitoring System (PMS), which tracks the user's maximum accumulated privacy disclosure probability in real time based on historically released information. Existing works mainly focus on static data, but daily life experiences tell us that some attribute values may change over time, and some initially sensitive information may become non-sensitive. Attackers may adopt fuzzy matches to handle the inconsistent data during the linkage attacks. We conduct extensive simulations to study the impacts of changeable sensitive and regular attribute values and fuzzy matches on privacy leakages and evaluate the performance of PMS in various situations.

A Privacy-Aware System Using Threat-Based Evaluation and Feedback Method in Untrusted Ubiquitous Environments.

A Threat-Based Privacy Preservation System in Untrusted Environment

A Novel Threat Evaluation Method for Privacy-Aware System in RFID

Enhanced User Data Privacy with Pay-by-data Model

Threat-based Evaluation for Context-Aware Multimedia Surveillance System.

A Privacy Protection Model of Data Publication Based on Game Theory

From Data Disclosure to Privacy Nudges: A Privacy-aware and User-centric Personal Data Management Framework

A Privacy Risk Assessment Model Based on TAPE Framework

Privacy Protection among Three Antithetic-Parties for Context-Aware Services

A Purpose-Based Privacy-Aware System Using Privacy Data Graph

Uncertain Privacy Decision about Access Personal Information in Pervasive Computing Environments

A Trust-Based Evaluation Model for Data Privacy Protection in Cloud Computing.

A System for Risk Assessment of Privacy Disclosure

A Privacy Protection Model for Transparent Computing System.

A Study on Context-aware Privacy Protection for Personal Information

A Real-Time Privacy Monitoring System for Multi-Tasks-Based Privacy-Disclosure in Crowdsourcing

A Privacy-aware Graph-based Access Control System for Healthcare Domain.

Privacy-Preserving Data Collection in Context-Aware Applications

Digital Privacy Under Attack: Challenges and Enablers

Cloud Service Evaluation Model Based on Trust and Privacy-Aware

Seeing Isn't Believing: QoE Evaluation for Privacy-Aware Users.