Search

By Fairness Management we refer to a wide variety of techniques that aim to improve the procedural or distributional fairness of ODS. This chapter begins with a consideration of Fairness Management for resource sharing in an ODS, and presents various application areas where such a problem needs to be solved. The discussion of these application areas will be accompanied by case studies that present various Fairness Management techniques. Most of these techniques can be described using the concepts of the theory of fairness introduced in section 2.3. The case studies have been selected so that they demonstrate various kinds of fair distribution problems. The first example of a fair scheduling in grids is a decentralized problem, in which a solution cannot be imposed on the participants. On the other hand, in the second example of a network dimensioning problem, the fair solution can be implemented by a central control – network management. The two cases differ also in other respects; the first problem has an extremely large solution space and a heuristic method is presented for finding equitable solutions, while the second problem is formulated as a Mixed Integer Linear Optimization problem and solved by a CPLEX solver. As a third example, we consider fairness in Peer-to-Peer systems that are even more decentralized than grids. [ABSTRACT FROM AUTHOR]

Trust Management (TM) is an area of information technology that aims to improve the operation of open, distributed systems by predicting or influencing the behavior of their users. When applied to human users, Trust Management methods attempt to leverage the human capacity for trust or distrust. Alternatively, TM systems can also be applied to control the behavior of non-human agents. In such a case, TM algorithms can support the fully automated control of agent operations, and have an impact on emergent properties of the ODS. [ABSTRACT FROM AUTHOR]

Information technology today is a vast body of knowledge that has found applications in most areas of human activity and thought. With the proliferation of the Internet and its various applications, the role of information technology in society has become vital and is rapidly increasing in importance. It can be said that the Internet has become intertwined with society; and with each new generation of Internet users, information technology becomes applied not only more widely, but also affects our society and personal lives more deeply. We are witnessing the birth of the real e-society. [ABSTRACT FROM AUTHOR]

Internet auctions are used everyday by millions. However, despite frequent criticism, only the most simple reputation systems are used by the most popular Internet auctions today. As a consequence of this, an experienced auction user is forced to undergo the menial task of reading and judging comments about his potential transaction partners. While it is true that the human mind is the best possible method of evaluating this information, the task is time-consuming and error-prone: the sheer number of comments is sometimes an obstacle to making a good decision under uncertainty. An inexperienced auction user, on the other hand, is often daunted by the task of understanding the information provided to him. It takes some learning to understand the value of a negative feedback, to evaluate the contents of the comments, or to understand what it means that the auction system has returned the handling fee. The reason for this situation may be the fact that the management of auction sites uses other mechanisms, like auction insurance or escrow, to protect users against outright fraud. And, as has been argued by the management of auctions sites when we have had an opportunity to discuss the issue, the simplicity of the presently used reputation system is an added bonus: it creates the impression of a simple, easy-to-understand tool. The fact that this simple tool is vulnerable to several adversary strategies [1,2,3] and that its design has an adverse impact on the reporting behavior of users [4,5] is not a sufficient argument for a change. [ABSTRACT FROM AUTHOR]

Reducing Internet auction fraud is one of the greatest challenges in today΄s electronic market. Most of the electronic auction platforms use only simple reputation system that can be easily manipulated [1,2,3]. Although reputation systems can be used to detect frauds, they provide little detailed information about the fraud itself except user comments. [ABSTRACT FROM AUTHOR]

An auction platform is a dynamic environment where a rich variety of social effects can be observed. Most of those effects remain unnoticed or even hidden to ordinary users. The in-depth studies of such effects should allow us to identify and understand the key factors influencing users΄ behaviour. The material collected from the biggest Polish auction house has been analyzed. NLP algorithms were applied to extract sentiment-related content from collected comments. Emotional distance between negative, neutral and positive comments has been calculated. The obtained results confirm the existence of the spiral-of-hatred effect but also indicate that much more complex patterns of mutual relations between sellers and buyers exist. The last section contains a several suggestions which can prove useful to improve trustworthiness of users΄ reports and security of an auction platform in general. [ABSTRACT FROM AUTHOR]

Trust Management is widely used to support users in making decisions in open, distributed systems. If two sellers on e-Bay have similar goods and service, similar marketing, they should also have similar income and reputation. Such an expectation can be formulated as a hypothesis: in realistic reputation (or trust management) systems, fairness should be an emergent property. The notion of fairness can be precisely defined and investigated based on the theory of equity. In this paper, we investigate the Fairness Emergence hypothesis in reputation systems and prove that in realistic circumstances, the hypothesis is valid. However, Fairness Emergence is not a universal phenomenon: in some circumstances it would be possible for one of two similar sellers to be better off. We study the sensitivity of Fairness Emergence to various aspects of a reputation systems. [ABSTRACT FROM AUTHOR]