Skip to content

A secure co-owned data sharing framework using fuzzy group decision making and users reputation

Student thesis: Doctoral Thesis

The usage of online social networks (OSNs) has become a crucial main activity for individuals in recent days. In OSNs platforms, users are given a space where they can share various types of contents such as, photos, videos, texts, and events. Users are also allowed to share contents of data by including other users ids on the shared content. This type of data is called co-owned data in OSNs. The majority of privacy issues in OSNs platforms are caused by these types of data sharing. Users whose information is leaked, either choose to become unfriend with the user, who leak their privacy, or quit from OSNs platforms, both cases are contradictory to the main OSNs goal. There is a considerable amount of research work done in order to address the privacy issues and proposed solutions. However, privacy issues which originated from co-owned data sharing have still been a problem in OSNs. This research addresses privacy issues, originated from coowned data sharing processes in OSNs. For instance, users’ privacy is still being leaked in Facebook, which is one of the most popular social network, users therefore quit from Facebook or be unfriend with others for protecting themselves. Privacy leakage has a significant effects on people’ lives, such as loosing life, breaking up their relationships, be raped. Being unfriend or quitting from social networks are contradictory to main aim of online social networks. This research therefore introduces a framework which makes a balance between co-owned data sharing and privacy preservation.
The developed framework consisted of four main phases which are the contributions of it; (1) a fuzzy logic decision making system, (2) a group decision making system, (3) trust and reputation models, and (4) formal modelling of controlling flow of shared coowned content. To make these contributions of this theses, this research adopted two methodologies; the mathematical models are developed with adaptation of the scientific methodology and the build methodology is used to implement the developed models in a real world application. The quantitative study was used to model the equations in the developed framework.
In order to evaluate this thesis work, the work was evaluated with a critical comparison with similar works, and the implementation of the developed framework was evaluated with analysis on critical requirements. The main contribution of this thesis is a secure co-owned data sharing framework with mathematical models. The developed framework aims to make a balance between data sharing and privacy preserving in co-owned data sharing processes in OSNs.
The developed framework has provided the most secure co-owned data sharing process with its mathematical models and the systems which compromises the developed mathematical models. It has also shown that data sensitivity depends on the data security features, this means that in the all previous work data sensitivity value was either ignored or decided by someone. However, the person who decides the data sensitivity may not have any idea about it, therefore, this thesis data sensitivity mathematical model solves this issue. All the equations which are used in the developed framework are novel and robust. The robustness are tested based on the developed models’ behaviours. The novelty is that there is no mathematical models which can be used in a co-owned data sharing process. They are developed to make not only a trade off between co-owned data sharing and users privacy protection but also make co-owned data processes more secure. The comparison between the developed framework and the similar works in the area has shown that the trade-off between co-owned data sharing and users’ privacy protection is possible only if the proposed fuzzy group decision making systems and reputation models are used.
Original languageEnglish
Awarding Institution
Supervisors/Advisors
Award dateMay 2020
Relations Get citation (various referencing formats)

ID: 22365012