Context-aware security is a promising approach for overcoming many of the security problems in modern Internet networks and systems. Engineering context-aware security solutions for these challenges is difficult, since there is no precise understanding of the notion of context relevant for security. On the other hand such solutions are intended to be used in a number of different situations or use case. Through an analysis of diverse use cases ranging from social group based to content distribution oriented we have defined a conceptual model of security context. The model identifies the important concepts that constitute security context and the relations that exist between them. We show that our model is suitable for analysing target situations in advance from security perspective, as well as for representing the security context that state-of-the-art approaches take into consideration. Our model promises to facilitate the specification of security context and the management of context-aware security policies. The model paves the path for secure context-aware content distribution and consumption as perceived by the project.
COBISS.SI-ID: 27547431
Online trust systems aim to translate the role that trust has in the traditional world onto the virtual platforms based on the use of network infrastructure built with the modern P2P networks. Establishing the interdependence between these systems and the human factor is essential for reducing the inherent complexity of the open platforms, and for improving the user experience and system performance. They enable building trust and solid on-line reputation systems. This work determines the systemic features of trust and introduces a novel framework of design properties based on the principles of General Systems Theory. The systemic properties that were neglected in the current technical solutions were further studied, enriched with social and human factors into the design guidelines of the on-line trust and reputation systems.
COBISS.SI-ID: 21257190
The behaviour patterns resulting from the interactions of many trusting entities in ecommerce systems may be much more complex than the performance of each of the individuals separately; thus, simple rules of trusting behaviour give rise to complex, emergent patterns. A major reason these emergent properties were neither successfully captured nor adequately treated by the current formal models is the global trend of addressing issues related to technical systems in a mechanistic manner considering the system simply as a sum of its components and neglecting the interactions between those components. The work published introduces the concept of an organismic property of human centric ecommerce systems and reveals new areas of applicability of trust as an organismic system trait. The goal of the research presented, is twofold: providing a novel view of treating trust related issues in ecommerce systems, and pointing to the missteps that can be brought by a systemic ignorance of the organismic nature of online trust systems.
COBISS.SI-ID: 26036007
Trust and reputation comprise a wide research area in social sciences, but these are also pillars of many social phenomena that shape the Internet socio-economic scene especially in the user oriented services where users are looking for trustable recommendations and opinions. The blossoming of virtual communities, especially when social networks are considered, largely changed the way trust is on-line formed and propagated. The few existing taxonomies provide only initial insights into the ways trust-benefits can be felt; they are neither complete nor elaborated in a systemic manner to provide a proper framework guided by real system-principles. In this paper, we propose a multidimensional framework for guiding the design-process, and assessing the completeness and consistency of reputation systems. The framework developed in the presented study is based on System theory principles; it identifies reputation system components, and more importantly, defines their interrelations. It considers the interaction-centric, dynamic and environment dependent trust-establishment and detects five major factors that guide reputation mechanisms design. The presented framework was applied to BarterCast reputation mechanism deployed in the BitTorrent protocol based client known as Tribler. The published paper based on this study was first presented as an invited paper on the ServiceWave conference in 2011.
COBISS.SI-ID: 25130535
The future content delivery platforms are predicted to be efficient, user-centric, low-cost and participatory systems, with social and collaborative connotation. The peer-to-peer (P2P) architectures, especially ones based on BitTorrent protocol, give a solid basis for provision of such future systems. However, current BitTorrent P2P networks lack flexible access control mechanisms. In the paper enhancements regarding the efficiency of the existing access control mechanism for BitTorrent systems – the Closed Swarms protocol are presented, providing additional flexibility in access control mechanism, enabling fine grained security policies specification and enforcement. The enhancements fulfil a number of content providers’ requirements and promise efficient, flexible and secure content delivery in the future content delivery scenarios based on the P2P based protocols.
COBISS.SI-ID: 24977959