Community-based service location

Proceedings of the Second International Joint Conference on Autonomous Agents and Multiagent Systems

2003

Self Cite

Agents must decide with whom to interact, which is nontrivial when no central directories are available. A classical decentralized approach is referral systems, where agents adaptively give referrals to one another. We study the emergent properties of referral systems, especially those dealing with their quality, efficiency, and structure. Our key findings are (1) pathological graph structures can emerge due to some neighbor selection policies and (2) if these are avoided, quality and efficiency depend on referral policies. Further, authorities emerge automatically and the extent of their relative authoritativeness depends on the policies.

Section: Evaluation Metricsmentioning

confidence: 99%

Emergent properties of referral systems

Yolum

Proceedings of the Second International Joint Conference on Autonomous Agents and Multiagent Systems

2003

Self Cite

“…We can model a referral network as a directed graph each of whose nodes represents an agent and each of whose edges represents an agent (at the origin) having another agent (at the target) as a neighbor [3]. Each agent's expertise describes what knowledge it possesses and its interest determines what knowledge it seeks.…”

Section: Technical Framework and Definitionsmentioning

confidence: 99%

Emerging Properties of Knowledge Sharing Referral Networks: Considerations of Effectiveness and Fairness

Manavalan

Lecture Notes in Computer Science

2012

Self Cite

Abstract.Referral-based peer-to-peer networks have a wide range of applications. They provide a natural framework in which agents can help each other. This paper studies the trade-off between social welfare and fairness in referral networks. The traditional, naive mechanism yields high social welfare but at the cost of some agents-in particular, the "best" ones-being exploited. Autonomous agents would obviously not participate in such networks. An obvious mechanism such as reciprocity improves fairness but substantially lowers welfare. A more general incentive mechanism yields high fairness with only a small loss in welfare. This paper considers substructures of the network that emerge and cause the above outcomes.

“…One way of representing the set of services is through a vector space model, where each element in the vector corresponds to a different domain and the weight of the element denotes the fitness of the service for that domain [14]. This is similar to the vector descriptions of documents in information retrieval.…”

Section: Graph-based Representationmentioning

confidence: 99%

“…That is, in order to create trust through second-hand information, the trustworthiness of the information sources must be established as well [15, p. 74]. A powerful way of ensuring that the sources themselves are trustworthy is by accessing them through referrals [14]. Local trust means considering previous direct interactions with a trustee, which often are the most valuable in creating trust for the following reasons.…”

Section: Introductionmentioning

confidence: 99%

Service Graphs for Building Trust

Yolum

On the Move to Meaningful Internet Systems 2004: CoopIS, DOA, and ODBASE

2004

Self Cite

Abstract. Information systems must establish trust to cooperate effectively in open environments. We are developing an agent-based approach for establishing trust, where information systems are modeled as agents that provide and consume services. Agents can help each other find trustworthy parties by providing referrals to those that they trust. We propose a graph-based representation of services for modeling the trustworthiness of agents. This representation captures natural relationships among service domains and provides a simple means to accommodate the accrual of trust placed in a given party. When interpreted as a lattice, it enables less important services (e.g., low-value transactions) to be used as gates to more important services (e.g., high-value transactions). We first show that, where applicable, this approach yields superior efficiency (needs fewer messages) and effectiveness (finds more providers) than a vector representation that does not capture the relationships between services. Next, we study trade-offs between various factors that affect the performance of this approach.