Content Trust Evaluation for Web Applications
Based around the advancement of privacy protection and helping individuals to regain control over their personal data, there have been efforts and initiatives towards the redecentralization of the web. To deal with the large amount of data providers in decentralized systems, it will be necessary to use web applications that automatically select a sufficiently trustworthy source of information for the given task. This requires those web applications to make trust-aware decisions, which, in turn, requires a framework to adequately evaluate and judge trust automatically. An already existing approach for this trust judgement is called entity trust. Entity trust works by combining data about the reputation of an information source with authentication technologies. If the information source is being trusted, the information will be trusted as well. Consequentially, the result of a trust evaluation using entity trust is absolute and a comparison of trust between two sources of information is not possible. Furthermore, this form of trust judgement does not consider the context or content of the evaluated information. Therefore, content trust is possibly a better approach that considers data about content and context of the evaluated information to create a basis for more nuanced trust decisions. Content trust realizes this by separating into 19 influencing factors to calculate a final trust value.
The objective of this thesis is to develop and implement a framework that allows web applications to independently evaluate the trustworthiness of an information by using a content trust centered approach. Instead of a simple dichotomous answer to the question whether the information should be trusted or distrusted, the framework will produce a value that enables a more complex and nuanced trust-aware decision by the web application. This thesis solely focuses on a first implementation prototype for the trust evaluation of the 19 factors of content trust. The trust scale used by the framework should be exchangeable and tested with the scale from Marsh and Briggs. Further, the framework design requires to evaluate all 19 different factors based on specific input information per factor. These can be given by expert knowledge and described for evaluations with the aTLAS scenarios, as the input information creation at runtime is not part of this thesis. The characteristics and details of trust-aware decisions will not be discussed as well.
The objective of this thesis is to find an approach or combination of approaches for the previously mentioned problems and tasks in the context of content trust for web applications in a redecentralized web. This particularly includes the state of the art regarding trust frameworks for web applications. The demonstration of feasibility with an implementation prototype of the concept is part of this thesis as well as a suitable evaluation with exemplary use cases.