On the Reputation of Agent-Based Web Services.

Conference Paper (PDF Available) · January 2010with29 Reads
Source: DBLP
Conference: Proceedings of the Twenty-Fourth AAAI Conference on Artificial Intelligence, AAAI 2010, Atlanta, Georgia, USA, July 11-15, 2010
Maintaining a sound reputation mechanism requires a ro bust control and investigation. In this paper, we propose a game-theoretic analysis of a reputation mechanism that objectively maintains accurate reputation evaluation of selfish agent-based web services. In this framework, web services are ranked using their reputation as a result of provided feed back reflecting consumers' satisfaction about the of fered ser vices. However, selfish web services may alter their public reputation level by managing to get fake feedback. In this paper, game-theoretic analysis investigates the payof fs of dif ferent situations and elaborates on the facts that discourage web services to act maliciously. Copyright © 2010, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.


    • "The main aim of this paper is to show the efficiency of the cuckoo search algorithm [5] in solving UCP and to propose a fresh MOCS to reject the role of the conclusion maker (system operator) in setting the reeling reserve in UCP. In MOUCP problem [12], is similar to other evolutionary process, cuckoo search algorithm begins with an initial frozen number of cuckoo nests [7]. At the end of every propagation of CSA, the size of cuckoo nests gains when compared to the sign number of nests. "
    [Show abstract] [Hide abstract] ABSTRACT: Web Service is an open standard based web applications that interact with other web applications for the reason of exchanging data. Web service is self telling service that will perform fine tasks and can be accessed through the web. Web services discovery have the Hyperlink-Induced Topic Search (HITS) linking analysis algorithm that rates Web pages. The increasing convolution of compositions and the ration of people and services involve adaptive and context-aware contact models. This method is an imperative for the internet penetrating techniques. But this emerging technique faces Topic drift problem which is a pitfall in web service discovery. This paper we introduce HYBRID algorithm to endorse Topic drift drawback. This algorithm is combination of Multi Objective Bat Algorithm (MOBA) and Multi objective cuckoo search (MOCS) algorithms .MOBA is used for the multi data access and the MOCS algorithm is used for fast access techniques .By using this HYBRID algorithm the efficiency and accuracy can be improved. So this technique is called as HYBRID algorithm (MOBA &MOCS). The experimental evaluation shows that this algorithm reduces topic drifting problem and increases the accuracy of the web services discovery.
    Full-text · Article · Jan 2015 · Expert Systems with Applications
    • "Thus, buying agents need a means to assess the quality of different sellers offering a particular product and select the most profitable seller who best meets the buyers' requirements. Reputation systems [8,9,11,16] are particularly effective approaches for buyers to evaluate sellers. In such systems, if a certain buyer does not have adequate personal experience of some candidate sellers, he will disseminate queries to other buyers (also called advisers) to request information about the sellers. "
    [Show abstract] [Hide abstract] ABSTRACT: In competitive electronic marketplaces where some selling agents may be dishonest and quality products offered by good sellers are limited, selecting the most profitable sellers as transaction partners is challenging, especially when buying agents lack personal experience with sellers. Reputation systems help buyers to select sellers by aggregating seller information reported by other buyers (called advisers). However, in such competitive marketplaces, buyers may also be concerned about the possibility of losing business opportunities with good sellers if they report truthful seller information. In this paper, we propose a trust-oriented mechanism built on a game theoretic basis for buyers to: (1) determine an optimal seller reporting strategy, by modeling the trustworthiness (competency and willingness) of advisers in reporting seller information; (2) discover sellers who maximize their profit by modeling the trustworthiness of sellers and considering the buyers’ preferences on product quality. Experimental results confirm that competitive marketplaces operating with our mechanism lead to better profit for buyers and create incentives for seller honesty.
    Full-text · Article · Nov 2014
    • "Without a reputation enabling mechanism, users cannot differentiate among services, specially the ones which offer the same type of service. Reputation mechanisms usually aggregate users' experiences (Khosravifar, Bentahar, Moazin, & Thiran, 2010b ), and in our case it strongly depends on QoS that each service provides. Users define tasks, each one with specific quality T r QoS , so that after performing a certain number of tasks, each one with QoS r w , during a window time t, the reputation of w gets evaluated by the master agent. "
    [Show abstract] [Hide abstract] ABSTRACT: Frameworks for aggregating similar services into structures called communities have been recently advocated. A common assumption in those frameworks is that residing services are coopetitive, i.e., competing over received requests, but also cooperating, for instance in terms of substituting each other. In this coopetition context, deciding to compete or cooperate at different moments in time is an open question yet to be addressed. The contribution of this paper is the answer to this challenging question by proposing a game-theoretic-based decision mechanism that services can use to effectively choose competition or cooperation strategies that maximize their payoffs. To achieve this objective, we investigate autonomous services’ characteristics and their expected utilities over different strategies. We propose a game-theoretic best response technique to measure the threshold that services can use in order to decide about the two strategies. We prove that the proposed decision mechanism is efficient and can be implemented in time linear in the length of the time period considered for the analysis and the number of services in the community. Moreover, we conduct extensive simulations to analyze various scenarios and confirm the obtained theoretical results. Those results show that our model outperforms existing competitive and random coopetitive strategies and the more services deviate from our game-theoretic-based coopetitive strategy the more they make less benefits.
    Full-text · Article · Aug 2014
Show more