Διδακτορικές διατριβές

Μόνιμο URI για αυτήν τη συλλογήhttps://pyxida.aueb.gr/handle/123456789/53

Περιήγηση

A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω

Τώρα δείχνει 1 - 1 από 1

Managing the risk of peer-to-peer transactions
Androutsellis-Theotokis, Stephanos; Ανδρουτσέλλης-Θεοτόκης, Στέφανος; Athens University of Economics and Business, Department of Management Science and Technology; Spinellis, Diomidis; Doukidis, Georgios; Louridas, Panos
It is being progressively recognized that information systems and applications supporting collaborative tasks, including online transaction processing systems, that currently follow centralized client-server models, can also be based on the maturing wave of peer-to-peer architectures. Motivations for this architectural decision include improved scalability, performance, and access to resources; ability to deal with transient user populations; dynamic ad-hoc communication, network organization and failure recovery; lack of a single point of failure; and ability to perform direct and unmediated transactions. In order to manage and reduce the risk inherent in peer-to-peer transactions and their decentralized and uncontrolled environment, a variety of approaches have been proposed in the literature and implemented in both research and industrial settings, with reputation and trust management systems being the most prominent. These aim to provide peers with estimates of the risk involved in their transactions, based on the observed past behaviour of their counterparties. Though reputation management systems offer a lot in this direction, it is argued that the information they provide about past behaviour may not be enough to accurately assess the risk involved in a transaction, and various issues remain open especially in purely distributed implementations. As a result, there still is considerable concern over the amount of risk involved in online transactions, even with the support of reputation management systems. In our research we attempt to address the above issues by proposing new, alternative and original concepts for reducing the risk of peer-to-peer transactions, and by designing, implementing and evaluating decentralized risk management systems that incorporate and utilize these concepts. After investigating the emerging field of peer-to-peer systems and architectures, and distributed reputation management systems, we first focus on the way in which reputation information is currently modeled, expressed and used, as it is an area that we believe warrants further research. In this work we propose an alternative approach, in which reputation information is expressed and manipulated in terms of monetary units. By coupling the reputation information with the transaction amount we can describe and implement specific algorithms for the estimation and propagation of reputation information, and other decision making processes. Central to this approach is the notion of “ratcheting” trust estimates, i.e. allowing the build-up of trust as a result of repeated successful transactions, potentially beyond the actual transaction value. Based on these concepts we design, test and evaluate the MOR-TRUST distributed reputation management system. We then take the above work a step further, to propose an alternative concept for peer-to-peer transaction risk management, and a corresponding system that implements it. In the PTRIM system we no longer rely on a distributed reputation management approach, but instead we base our design on the financial principles governing credit markets for managing, transferring or reducing credit and transaction risk. PTRIM builds a transaction default market layer on top of a main transaction processing system, within which peers offer to underwrite the transaction risk for a slight increase in the transaction cost. The insurance cost, determined through market-based mechanisms, is a way of identifying untrustworthy peers and perilous transactions. The risk of the transactions is contained, and at the same time members of the peer-to-peer network capitalize on their market knowledge by profiting as transaction insurers. Although PTRIM extends the main concepts that the MOR-TRUST approach is based on, the two system designs and implementations are independent. We define a robust evaluation methodology for each of the two systems. For MORTRUST we build a simulation environment based on synthetic network graphs and data, and run sets of simulation trials. For PTRIM, instead, we design, implement and deploy a complete prototype system and run sets of trials based on groups of real subjects, as well as a different “composite” platform involving real transaction data sampled from actual online transaction processing systems and real subjects providing insurance for these transactions. The evaluation of these two systems validated and verified the efficiency of our proposed concepts. Our findings suggest that our proposed approaches are able to support peer-to-peer transaction processing systems either through distributed reputation management, or through the insurance offers produced by the transaction default market layer. At the same time, we highlight areas of future work and possible improvements with respect to both the systems and the evaluation procedure itself.

Περιήγηση

Πλοήγηση Διδακτορικές διατριβές ανά Συγγραφέα "Androutsellis-Theotokis, Stephanos"