Back to search

University of the Aegean

Enabling privacy-preserving SNS communications by design: a stakeholders’ approach

Abstract

dc:description

SNS users communicate via SNS platforms, sharing vast amounts of data in a semi-public medium and expect their privacy to be protected. Privacy studies place the responsibility in the SNS platform provider, usually an organization profiting from exploiting users’ data or focus on the exchanged message, while other aspects of the communication are left unprotected. At the same time, users’ privacy is violated by other users’ publications or their own actions exposing personal data. Thus, there is a need to protect users’ personal information taking into account all stakeholders’ responsibilities to protect SNS users’ privacy, in all aspects of communications taking place (including the message and the social network interactions). Analysis of existing literature revealed that although several privacy requirements elicitation methods have been proposed (threat-based, goal-driven approaches, mainly as extensions to security requirements elicitation methods), these do not take into account the complex relationships formed in SNS communications and do not consider all types of personal data to be protected. Also, the SNS paradigm, in which a new form of public-private sphere instantiates, calls for implementation of special privacy requirements. This Thesis’ objective is to guide privacy-preserving SNS interactions by design. To achieve this, it provides answers two research questions: a) which are the privacy requirements that need to be addressed for privacy-preserving SNS communications, considering all stakeholders and b) how can a platform implement these privacy requirements by design. To achieve the first research objective, a threat-based analysis of SNS communications was applied, in order to elicit privacy requirements. An in-depth analysis of literature on SNS, combined with the experimental analysis of popular SNS platforms allowed the identification of stakeholders involved in SNS communication and the different types of formed relationships among them (one-to-one, one-to-many, many-to-many), as well as the types of personal data explicitly or implicitly shared during different use case scenarios of SNS. Modelling of the SNS functionality in terms of communication models was then performed, based on this identification. The different models served as the basis for privacy threat analysis of the interactions of SNS users, in which all stakeholders were considered as a potential risk source for misusing personal data. Identification of privacy risks led to identification of measures to mitigate them, which form the privacy requirements. The result is a list of privacy requirements SNS platform providers, users and third-parties should implement to ensure communications in SNS are conducted in a privacy-preserving way. The privacy threat analysis also highlighted the significance of promoting awareness among SNS users and the role of user in protecting privacy by employing PETs. With a focus on promoting privacy by design and by default (to address the 2nd research question of the Thesis), we identified further privacy requirements by extending privacy-by-design strategies. This highlighted the significance of ensuring usability and user awareness, while implementing the privacy requirements.Further analysis of privacy awareness practices, by reviewing literature and evaluating their application in popular SNS platforms, indicated that there is need for SNS platforms to promote privacy practices enhancing the usability aspect of privacy settings (e.g. visualising risks, providing awareness functionality under organized privacy settings menus, differentiating from account security settings), as well as the content of notifications to the users, in order to assist them in understanding source of the potential risks in their shared personal data. In addition, we explored why users do not adopt privacy-enhancing tools, either embedded in SNS platforms or stand-alone. Our findings revealed that apart from awareness, users’ adoption of PETs is influenced by direct and indirect costs (e.g. for learning to use a tool or experiencing delays in the service), required technical expertise and feedback of its effectiveness. Especially for stand-alone PETs, a survey among 170 SNS users found that quality (effectiveness in protection), effectiveness feedback and low cost are the most important factors users weight in their decision to adopt a tool. These findings provide an insight on how embedded PETs could be presented to users (another requirement for SNS platform providers), but also explore the factors influencing adoption of stand-alone PETs and can serve as the basis for designing successful privacy-enhancing tools for SNS.Furthermore, in order to guide implementation of privacy by design, this Thesis explored the implementation of Privacy Impact Assessments (PIAs) (Oetzel and Spiekermann, 2014) and extended existing guidelines, as in practice, PIA conduction is still a difficult task for PIA practitioners. To answer the second research question of this Thesis and to provide practical guidance to PIA practitioners, and in extension to SNS designers acting on behalf of the SNS platform provider, we developed a framework to evaluate guidance provided by existing methods. Its application in 9 popular PIA methods revealed that guidance for identification of risks in many methods takes the form of compliance checklists of general legal requirements and could even mislead PIA practitioners and limit their view on privacy risks emerging from the personal data processing. Findings also included contradicting guidelines about organizing PIA projects and inadequate guidance on involving stakeholders in the process. In order to propose a set of comprehensive guidelines, we identified best practices of PIA methods and researched existing literature for areas not adequately described in PIA methods (e.g. assessing privacy risks). Furthermore, in this Thesis we propose an organizational scheme for PIA projects, guiding PIA practitioners in assigning roles and responsibilities in the context of a PIA (e.g. PIA conduction, signing-off the PIA report, implementing selected controls, ownership of residual risks, ownership of the PIA report and its updates whenever necessary and acting as an external contact point for audit authorities). Finally, to link theory with practice, this Thesis combines the findings of the above steps to a comprehensive PIA process, extending general guidelines provided in ISO 29134. The main contribution of this Thesis is a framework for privacy by-design in SNS, which identifies and guides implementation of privacy requirements that need to be addressed, considering all stakeholders. This framework can also be applied in relevant types of online services where personal communications take place and private and public spheres co-exist and consists of three pillars: a) privacy analysis of the SNS communications paradigm through communication models, to reveal which entities take part in the communication and which types of personal data need to be protected b) elicitation of privacy requirements for SNS, by a privacy threat-based analysis and c) guidelines on Privacy Impact Assessment conduction. Our approach considers all stakeholders as responsible for privacy protection and proposes privacy requirements for the SNS platform, users and third-parties (research question 1). We also propose use of such an approach in the context of a PIA and provide comprehensive guidelines for PIA conduction, including an organizational scheme and an evaluation framework for PIA methods (research question 2). This Thesis contributes to privacy research in SNS and to privacy practices by:a)proposing an approach in which all stakeholders are considered responsible for privacy protection. Privacy analysis is facilitated by modelling SNS functionality as communications and extending the concept of message to include all types of shared personal data. b) proposing a privacy threat-based approach for privacy requirements elicitation in environments where private and public spheres co-exist,c)proposing a comprehensive list of privacy requirements, applying to all SNS communication models, organized per stakeholders, d) providing guidelines for SNS platforms to further adopt awareness practices, thus extending privacy requirements addressed to SNS platform providerse) providing a framework for users to evaluate effectiveness of proposed tools by the SNS platforms f) exploring privacy behaviour aspects, such as factors influencing the adoption of PETs, providing ground for further research and analysis to guide successful PETs design,g)providing a framework to evaluate proposed PIA methods. This can be used by PIA practitioners to decide whether a method provides them with sufficient guidance, based on their experience and the legal environment of their project, as well as by DPAs and researchers proposing new methods, h) proposing an organizational scheme describing PIA roles and responsibilities assignment and highlighting the main interactions among them, to guide PIA practitioners in organizing PIA projects and involving necessary stakeholders. i) providing practical guidelines for PIA practitioners extending ISO 29134, organized in a comprehensive PIA method. This method guides PIA practitioners through all steps of PIA conduction and highlights areas in need of more research such as metrics for privacy risk likelihood and impact assessment. Finally, as the interactions of SNS users involve sharing high volumes of personal data, we conclude that there is a continuous need to explore all stakeholders and their roles in emerging SNS models, in order to ensure privacy by design. Also, this Thesis identifies the need for further research in privacy-enhancing technologies from the technology and the human aspect, to ensure development of tools that offer adequate protection and are attractive to SNS users, forming privacy patterns, to assist designers in requirements implementation, as well as tools supporting and automating Privacy Impact Assessments.

Degree

thesis:*
Grantor dc:publisher
University of the Aegean
Year dc:date
2020

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Vemou, Konstantina
  • Βέμου, Κωνσταντίνα

Subjects

dc:subject × 12

Rights

Language dc:language
eng

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/48992

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Vemou, Konstantina; Βέμου, Κωνσταντίνα. Enabling privacy-preserving SNS communications by design: a stakeholders’ approach. University of the Aegean, 2020. http://hdl.handle.net/10442/hedi/48992