Fig 3 - uploaded by Maria Ganzha
Content may be subject to copyright.
Source publication
Agent technology is often claimed to be the most natural approach for automating e-commerce business processes. Despite these
claims, up till now, the most successful e-commerce systems are still based on humans to make the most important decisions
in various stages of an e-commerce transaction. Consequently, it is difficult to find successful actu...
Contexts in source publication
Context 1
... There is one Main container that hosts the CIC agent. Users (customers and merchants) can create as many containers they need to hold their Client and Shop agents (e.g. one container for each e-store). Buyer agents created by Client agents use JADE mobile agent technology to migrate to the Shop agent containers to engage in nego- tiations. In this context, a container simulates a marketplace where various Seller and Buyer agents meet and negotiate. Moreover, all these containers linked via the agent platform simulate a bazaar filled with marketplaces filled with trading agents. The current implementation is based on several Java classes organized into several categories. Each category is implemented as a separate Java package. – Agent classes . Classes of this package are used for describing various agent types used in the system. Each agent class incorporates a subset of agent activity classes, also called behaviors. Behaviors are used as an abstraction that represents an atomic activity performed by an agent. – Database classes . Classes of this package are used for describing agents that are responsible for management of database connections. – Negotiation classes . Classes of this package implement a simple framework for de- scribing various negotiation protocols. This framework uses the Initiator and Par- ticipant roles, as defined by the FIPA Contract Net Interaction protocol ([5]). – Reasoning classes . These classes used for the implementation of the various reason- ing models employed by the negotiation agents; see [11] for more details concern- ing model of negotiation agents. Our implementation supports agents that dynami- cally load their negotiation protocols and reasoning modules. The implementation combines the Factory design pattern ([4]) and dynamical loading of Java classes ([11]). – Ontology classes . These classes are necessary for implementing agent communi- cation semantics, using concepts and relations. Current implementation uses an extremely simple ontology that defines a single concept for describing Client and Shop preferences including prices, product names and negotiation protocols. – Other classes . This package contains various helper classes. In our system, agent communication is implemented using FIPA ACL messages [5]. We have used the following messages: SUBSCRIBE, REQUEST, INFORM, FAIL- URE, CFP, PROPOSE, ACCEPT-PROPOSAL, REJECT-PROPOSAL, REFUSE. SUBSCRIBE messages are used by the Shop and Client agents to register with the CIC agent and for the Buyer agents to register (to participate in auctions) with the Seller agent. REQUEST messages are used by Client agents to query the CIC agent about what shops are selling a specific product and for Client agents to ask the Shop agent for a final confirmation of a transaction. INFORM messages are used as re- sponses to SUBSCRIBE or REQUEST messages. For example, after subscribing to the CIC agent, a Client agent will get an INFORM message that contains its ID, or after requesting the names of the shops that sell a specific product, a Client agent will receive a list of the Shop agent IDs in an INFORM message. Buyer agents are using FAILURE messages to inform the master Client agents about the unsuccess- ful result of an auction. Finally, CFP, PROPOSE, ACCEPT-PROPOSAL, REJECT- PROPOSAL and REFUSE messages are being used by negotiating agents. The system can be run in a simple setting for demonstration purposes by manually creating Shop and Client agents via the GUI, or directly from command-line when a large number of agents, containers, products etc. is to be created [6]. For the purpose of this paper we have utilized experiments involving multiple agents residing on multiple computers. First, Client agents resided on a single computer and Buyers migrated to Shop agents residing on the remaining 19 machines. Second, Client agents resided on 4 computers, while the remaining 16 machines contained Shop agents. Furthermore, to illustrate heterogeneity of the environment in which our system can run, in both experimental settings the Main container of the agent platform resided on a computer running Linux, while the remaining 20 computers run Windows. In addi- tion JADEs Sni ff er agent also was executed, on the Linux PC,. This agent is provided by JADE and its role is to report on communications between agents in the system. Figure 2 presents agent communication captured with help of this agent (note Linux environment). In the experiment shown in Fig 2 every Shop had three di ff erent products. Thus, at the beginning of an experimental run every Shop registered with the CIC agent , then created 3 Sellers (one Seller for each product). Seller agents also registered with the CIC agent and then waited for the incoming Buyer(s) . Communication involved in these operations can be seen in Fig 2. There exist two events which are necessary for to start negotiations: appearance of at least one Buyer and an interrupt caused by the timer (see Figure 3). After creation, Client registered with the CIC agent . Upon user request, it obtained list of Shops , where product(s) of interest were sold and created Buyer agents and sends them to the selected Shops . When Buyer arrived at the marketplace it asked about cur- rent negotiation protocol, communicated with its Client and obtained a corresponding strategy module and waited for start of negotiations. After finishing negotiations, Seller informed Shop agent about their results and Shop agent notified appropriate Client about successful result of negotiations (see also Fig 2). In the experiment represented in Fig 2 and Fig 3 we used three products, which Client could buy. Thus, we had a total of more than 200 agents populating the system. It should be pointed out that the most time-consuming operation is system initialization (creation of containers). However, since containers are created once, they have only minimal impact on the operations of the system. We have run multiple experiments, changing the number of (a) containers, (b) com- puters, (c) Clients , (d) Shops , (e) negotiation protocols, (f) products (g) mixture of Linux and Windows environments, etc. In each case experiments run smoothly and supported our general claim that the proposed system, when further developed can: (1) can be scaled to a truly large size, and (2) be used for e-commerce modeling. In this paper we have introduced an agent-based e-commerce system that has actually been implemented and show to fulfill the basic promises of agent systems. The most important of them were: (1) system scalability, (2) flexibility, and (3) heterogeneity. Obviously, the proposed system has a number of shortcomings that we are aware o ff , and we will work vigorously to remove them and develop and implement a truly com- prehensive system. We will report on our progress in subsequent ...
Context 2
... Java classes organized into several categories. Each category is implemented as a separate Java package. – Agent classes . Classes of this package are used for describing various agent types used in the system. Each agent class incorporates a subset of agent activity classes, also called behaviors. Behaviors are used as an abstraction that represents an atomic activity performed by an agent. – Database classes . Classes of this package are used for describing agents that are responsible for management of database connections. – Negotiation classes . Classes of this package implement a simple framework for de- scribing various negotiation protocols. This framework uses the Initiator and Par- ticipant roles, as defined by the FIPA Contract Net Interaction protocol ([5]). – Reasoning classes . These classes used for the implementation of the various reason- ing models employed by the negotiation agents; see [11] for more details concern- ing model of negotiation agents. Our implementation supports agents that dynami- cally load their negotiation protocols and reasoning modules. The implementation combines the Factory design pattern ([4]) and dynamical loading of Java classes ([11]). – Ontology classes . These classes are necessary for implementing agent communi- cation semantics, using concepts and relations. Current implementation uses an extremely simple ontology that defines a single concept for describing Client and Shop preferences including prices, product names and negotiation protocols. – Other classes . This package contains various helper classes. In our system, agent communication is implemented using FIPA ACL messages [5]. We have used the following messages: SUBSCRIBE, REQUEST, INFORM, FAIL- URE, CFP, PROPOSE, ACCEPT-PROPOSAL, REJECT-PROPOSAL, REFUSE. SUBSCRIBE messages are used by the Shop and Client agents to register with the CIC agent and for the Buyer agents to register (to participate in auctions) with the Seller agent. REQUEST messages are used by Client agents to query the CIC agent about what shops are selling a specific product and for Client agents to ask the Shop agent for a final confirmation of a transaction. INFORM messages are used as re- sponses to SUBSCRIBE or REQUEST messages. For example, after subscribing to the CIC agent, a Client agent will get an INFORM message that contains its ID, or after requesting the names of the shops that sell a specific product, a Client agent will receive a list of the Shop agent IDs in an INFORM message. Buyer agents are using FAILURE messages to inform the master Client agents about the unsuccess- ful result of an auction. Finally, CFP, PROPOSE, ACCEPT-PROPOSAL, REJECT- PROPOSAL and REFUSE messages are being used by negotiating agents. The system can be run in a simple setting for demonstration purposes by manually creating Shop and Client agents via the GUI, or directly from command-line when a large number of agents, containers, products etc. is to be created [6]. For the purpose of this paper we have utilized experiments involving multiple agents residing on multiple computers. First, Client agents resided on a single computer and Buyers migrated to Shop agents residing on the remaining 19 machines. Second, Client agents resided on 4 computers, while the remaining 16 machines contained Shop agents. Furthermore, to illustrate heterogeneity of the environment in which our system can run, in both experimental settings the Main container of the agent platform resided on a computer running Linux, while the remaining 20 computers run Windows. In addi- tion JADEs Sni ff er agent also was executed, on the Linux PC,. This agent is provided by JADE and its role is to report on communications between agents in the system. Figure 2 presents agent communication captured with help of this agent (note Linux environment). In the experiment shown in Fig 2 every Shop had three di ff erent products. Thus, at the beginning of an experimental run every Shop registered with the CIC agent , then created 3 Sellers (one Seller for each product). Seller agents also registered with the CIC agent and then waited for the incoming Buyer(s) . Communication involved in these operations can be seen in Fig 2. There exist two events which are necessary for to start negotiations: appearance of at least one Buyer and an interrupt caused by the timer (see Figure 3). After creation, Client registered with the CIC agent . Upon user request, it obtained list of Shops , where product(s) of interest were sold and created Buyer agents and sends them to the selected Shops . When Buyer arrived at the marketplace it asked about cur- rent negotiation protocol, communicated with its Client and obtained a corresponding strategy module and waited for start of negotiations. After finishing negotiations, Seller informed Shop agent about their results and Shop agent notified appropriate Client about successful result of negotiations (see also Fig 2). In the experiment represented in Fig 2 and Fig 3 we used three products, which Client could buy. Thus, we had a total of more than 200 agents populating the system. It should be pointed out that the most time-consuming operation is system initialization (creation of containers). However, since containers are created once, they have only minimal impact on the operations of the system. We have run multiple experiments, changing the number of (a) containers, (b) com- puters, (c) Clients , (d) Shops , (e) negotiation protocols, (f) products (g) mixture of Linux and Windows environments, etc. In each case experiments run smoothly and supported our general claim that the proposed system, when further developed can: (1) can be scaled to a truly large size, and (2) be used for e-commerce modeling. In this paper we have introduced an agent-based e-commerce system that has actually been implemented and show to fulfill the basic promises of agent systems. The most important of them were: (1) system scalability, (2) flexibility, and (3) heterogeneity. Obviously, the proposed system has a number of shortcomings that we are aware o ff , and we will work vigorously to remove them and develop and implement a truly com- prehensive system. We will report on our progress in subsequent ...
Citations
... Meeting, interaction and trading of buyers and seller is now facilitated either through direct contact, or via intermediation agents, with reduced cost and increased gain. Computer Scientists have developed digital decentralized markets where business partners assisted by software agents can register their capabilities, search for potential partners, and involve in trading activities [6,13]. ...
... Let us now consider the following cycle of : (1, 3), (3,6), (6,4), (4,1). Representing arcs by their indices and assigning one of signs + and − to each arc following its orientation relative to the cycle orientation, we get the cycle: 1, 5, −6, −2. ...
... Let us now consider the following cycle of : (1, 3), (3,6), (6,4), (4,1). Representing arcs by their indices and assigning one of signs + and − to each arc following its orientation relative to the cycle orientation, we get the cycle: 1, 5, −6, −2. ...
In this work we address the problem of optimizing collective profitability in semi-competitive intermediation networks defined as augmented directed acyclic graphs. Network participants are modeled as autonomous agents endowed with private utility functions. We introduce a mathematical optimization model for defining pricing strategies of network participants. We employ welfare economics aiming to maximize the Nash social welfare of the intermediation network. The paper contains mathematical results that theoretically prove the existence of such optimal strategies. We also discuss computational results that we obtained using a nonlinear convex numerical optimization package.
... El comercio electrónico comprende etapas o funciones que se deben cumplir durante su ciclo de vida, de las cuales podemos mencionar algunas como: administración de usuarios, administración de contenido, comercialización, negociación, administración de órdenes, procesamiento de pagos, servicio y asistencia [5][6][7][8]. ...
El comercio electrónico es un campo en constante evolución que involucra un con-junto complejo de redes, bases de datos, tecnologías de información y procesos. Con la presencia de internet, las empresas requieren aplicaciones que brinden se-guridad, administración confiable, integración, automatización y estandarización de procesos e información, además de mecanismos para apoyar la toma de de-cisiones y novedosas formas de presentar sus productos o servicios. Actualmente existen tecnologías, como los servicios web, los agentes y el modelado virtual, que soportan las necesidades requeridas, sin embargo trabajan de manera aislada. Con la finalidad de lograr procesos exitosos, automatizados, con manejo eficiente de in-formación y que conjunten las ventajas de cada tecnología, se propone un modelo de integración lógica de las mismas, mediante el cual se brinde a las empresas una solución integral para sus diferentes necesidades: operacionales, de administración eficiente de la información y de presentación novedosa de ésta en modelos de negocio B2B.
... The MANS was developed under JADE 3.0, which is fully implemented in Java language and is a framework for the development of multi-agent systems in compliance with FIPA specifications [1][2]. As a detailed development environment, we used FIPA ACL, JDK1.4, ...
Negotiation is a process of reaching an agreement on the terms of a transaction, such as price, quantity, for two or more parties. Negotiation tries to maximize the benefits for all parties concerned. Instead of using human-based negotiation, e-commerce provides such an environment as adopting automated negotiation. Thus, choosing agent technology is appropriate for an automatic electronic negotiation platform, since autonomous software agents strive for the best deal on behalf of the human participants. Negotiation agents need a clear-cut definition of negotiation models or strategies. In reality, most bargaining systems embody nearly one negotiation model. In this article, we present a mobile agent negotiation system with reusable negotiation strategies that allows agents to dynamically embody a user's favorite negotiation strategy which can be preinstalled as a component in the system. We develop a prototype system, which is fully implemented in compliance with FIPA specifications, and then, describe the benefits of using the system.
... For instance, one of important factors that influences the way that the SA interacts with incoming BAs is trust (see for instance [7, 28]). It should also be mentioned that in our system we utilize a modified negotiation framework [3, 4, 6] introduced originally by Bartollini, Jennings and Preist [8, 9]. In this framework, the negotiation process was divided into a generic negotiation protocol and a negotiation template that contains parameters of a given negotiation. ...
Large body of recent work has been devoted to multi-agent systems utilized in e-commerce; in particular, autonomous software agents participating in auctions. In this context we modify a model agent-based e-commerce system so that it can serve as an airline ticket auctioning system. Such a system can be then combined with a Travel Support System that utilizes ontologically demarcated travel-content. To achieve this goal, air travel data has to be demarcated utilizing an air travel ontology that has to support existing domain-specific real-world standards. One of such standards that steadily gains popularity in the air travel industry (and other travel areas) is the Open Travel Alliance (OTA) messaging system that defines, among others, the way that entities should communicate about air travel related issues. The aim of this paper is to outline our efforts leading toward creating an agent-based system for selling airline tickets that utilizes an air-travel ontology that matches the OTA messaging specification as well as satisfies procedures described in IATA manuals.
... The goal of developing and experimenting with a multiagent e-commerce system was recently specified in ([2] [10]). One of challenges is to illustrate that it is possible to use autonomous agents to completely substitute for humans to automate basic e-commerce activities such as: product brokering , merchant brokering, negotiations, payment etc., in a complete e-commerce scenario, rather than in isolation ([1]). ...
This note describes a sample implementation of automated negotiations in an e-commerce modeling multi-agent system. A specific set of rules is used for enforcing negotiation mech- anisms. Discussion of system design and implementation using JADE and JESS is provided. Finally, an experiment involving multiple English auctions performed in parallel is discussed.
With the advancements in World Wide Web and high speed internet the scenario of traditional shopping system has changed completely and e-shopping has gained popularity as it furnishes human with a feeling of easiness. But this e-shopping system is also being badly off for the involvement of human efforts and time to a great extent. To minimize this human intervention automation in e-commerce is like a blessing and use of agent technology is the best way to achieve this. In this context use of negotiation strategies and protocols mentioned in literature in developing systems which automates the whole e-shopping system is required. And so, here in this paper we present a multi-agent application using JADE, based on real world e-commerce; it automates an English auction by applying Kasbah negotiation strategies in order to minimize human intervention and ensures that the product is being sold at its best price.
Negotiation is a process of reaching an agreement on the terms of a transaction, such as price, quantity, for two or more parties. Negotiation tries to maximize the benefits for all parties concerned. Instead of using human-based negotiation, e-commerce provides such an environment as adopting automated negotiation. Thus, choosing agent technology is appropriate for an automatic electronic negotiation platform, since autonomous software agents strive for the best deal on behalf of the human participants. Negotiation agents need a clear-cut definition of negotiation models or strategies. In reality, most bargaining systems embody nearly one negotiation model. In this article, we present a mobile agent negotiation system with reusable negotiation strategies that allows agents to dynamically embody a user's favorite negotiation strategy which can be preinstalled as a component in the system. We develop a prototype system, which is fully implemented in compliance with FIPA specifications, and then, describe the benefits of using the system.
In this paper we present a framework for automation of distributed market place that is based on intelligent software agents and semantic web services. A multi-agent environment designed for automation of e-commerce activities such as negotiation between buyer and seller, recommendation, and customer decision making. Furthermore we use a semantic web service middleware called StraBM , advising managers of e-shops for making appropriate strategic decisions and managing business models. StraBM uses a customized WSMX middleware to generate appropriate strategy and business models.
As the mobile agent technology begins to emerge as a viable solution for intelligent e-commerce applications, there is an increasing attempt to ensure that the systems being developed are robust, reliable and manageable. While the mobile-agent-based e-commerce systems are desgined and implemented, their functional correctness and completeness must be formally verified. For this purpose, this paper introduces an extension of Petri nets, mobile-agent-oriented Petri nets (MAPN), for modeling and analyzing transaction workflows in mobile-agent-based e-commerce systems. By attaching some attributes to the tokens of the classical Petri nets, every token can be regarded as an agent to roam around the places to excutive certain tasks. In order to illustrate our formal model (MAPN) is effective for mobile agents modeling in e-commerce systems, an example of product search between buyers and sellers is provided. Meanwhile, the reach ability and the cover ability of the model are analyzed as well. Compared with exsiting formal description of agent-based systems, the tokens in MAPN model can carry more information, make the model more simple and powerful.
