Privacy for Peer Profiling in Collective Adaptive Systems

Abstract

Abstract. In this paper, we introduce a privacy-enhanced Peer Manager, which is a fundamental building block for the implementation of a privacypreserving collective adaptive systems computing platform. The Peer Manager is a user-centered identity management platform that keeps information owned by a user private and is built upon an attribute-based privacy policy. Furthermore, this paper explores the ethical, privacy and social values aspects of collective adaptive systems and their extensive capacity to transform lives. We discuss the privacy, social and ethical issues around profiles and present their legal privacy requirements from the European legislation perspective.

A Programming Model for Hybrid Collaborative Adaptive Systems

“…Manages all peer and application information and allows privacy-aware access and sharing of the peer/collective data among platform components and applications. Details on how the PeerManager is implemented can be found in [9]. Orchestration Manager (OM): Component in charge of orchestrating collaborative activities among human peers.…”

“…In order to fulfill the HDA-CAS design requirement of privacy compliance, the programming framework inherits the privacy features of the PeerManagr component described in [9]. It allows the programming framework to fetch peer profiles from the peerstore through a privacy-enforcing PM API containing only the data (possibly semantically obfuscated) previously approved by the user himself (upon signing up and creating the profile) for use by that particular application.…”

Ontology-Based Obfuscation and Anonymisation for Privacy

“…In [9], the authors introduce the concept of a privacy-enhanced Peer Manager, in which the original idea was designed to preserve privacy in collective adaptive systems. The Peer Manager works as an user-centered identity management platform that keeps user's information private.…”

“…The approaches discussed here employ obfuscation in different scenarios yet with similar goals. In summary, we aim to integrate the proposals [13] and [18] into the Peer Manager [9] obfuscation functions, and thus, demonstrate how it can be applied to healthcare systems (e.g., ehr and phr) using real medical ontologies.…”

“…The o-dss provides two fundamental functions: semantic obfuscation (so) and data anonymisation (da). The so function is specially grounded on the proposals: peer profiling [9] and context obfuscation [18]. That is, we aim to partially show or obfuscate the record (i.e., to provide data minimisation instead of anonymisation), based on the concept of Peer Manager for access control, and also, exploit the medical ontologies for data obfuscation.…”