Abstracts:

James Odell
James Odell Associates
3646 W. Huron River Drive
US - MI 48103, Ann Arbor, USA
E-mail: email@jamesodell.com

Leading-edge organizations are now developing systems that employ autonomous, interactive entities, or agents. [1; 2] Compared to its predecessors, the agent-based approach is evolutionary. However, its usages could be revolutionary. This paper begins by presenting some of the differences and similarities between agents and previous approaches. We then discuss some of the challenges for using current modeling approaches to represent agent-based systems. Our position is two folded: many of the evolutionary aspects of agent modeling can accomplished by extending current modeling languages such as UML 2.0; while the revolutionary aspects, however, will probably require new approaches. (pp. 3-11)

Keywords: systemic, agents, multi-agent system, UML, AUML, model, modeling languages, modeling notation

Sabin Buraga and Lenuţa Alboaie
Faculty of Computer Science, "A.I.Cuza" University of Iaşi, Romania
E-mail: busaco@infoiasi.ro, adria@infoiasi.ro
WWW: http://www.infoiasi.ro/~busaco/

In this paper, we present an agent-based object-oriented solution to access the Web distributed resources. We describe Omega – an agent framework viewed as a hierarchical space of a set of distributed objects that models the Web resources. Also, we propose an XML-based model that can be used as a universal manner for serialization of the objects processed by the (mobile) agents. The serialization mechanism can use the Simple Object Access Protocol (SOAP) serialization facilities, also. (pp. 13-22)

Hong Zhou, Yufang Wang, Dia Ali and Maria Cobb
Department of Computer Science and Statistics
University of Southern Mississippi
US - 39406-5601 Mississippi, Hattiesburg, USA
E-mail: hong.zhou@usm.edu

Shahram Rahimi
Department of Computer Science
Southern Illinois University
US - 62901Illinois, Carbondale, USA

E-mail: rahimi@cs.siu.edu

This paper presents a task-oriented compositional mobile agent architecture named AgentBee. In this architecture, the mobile agent is in fact a task component. Each task component is recursively formed from sub-task components together with information defining the relationships among sub-task components and supplementing data. At the lowest level of this composition there exist primitive task components which are conceptually indivisible. Such a task-oriented compositional structure allows easy modifications of mobile agents. Thus, not only can it facilitate knowledge exchange between mobile agents and agencies, but also it can enhance the assembly of new mobile agents from existing components. (pp. 23-30)

Sehl Mellouli, Bernard Moulin and Guy W. Mineau
Department of Computer Science and Software Engineering,
Laval University
CA - G1K 7P4 Quebec, Canada
E-mail: sehl.mellouli@ift.ulaval.ca, bernard.moulin@ift.ulaval.ca, guy.mineau@ift.ulaval.ca

Multi-Agent Systems (MAS) can be applied to a wide range of applications such as organizational, Internet, adaptive, or FIPA compliant applications. Developing such systems requires agent software engineering methodologies. To this end, many methodologies have been proposed such as Gaia (for organizational or Internet applications), ADELF (for adaptive applications) or SABPO (for FIPA compliant applications). However, no methodology has been proposed to deal with fault-tolerant MAS. In fact, agents are prone to failures, and thus are MASs. So, the MAS may not reach its objectives in case of agent failure. Hence, it is important to check a MAS design in order to prevent agent failures, so that we reduce failure risk at run time. The aim of this paper is to present an overview of agent oriented software engineering methodologies and to propose some guidelines towards a modelling methodology for fault-tolerant MAS. (pp. 31-40)

Keywords: methodology, multi-agent systems, design

Shahram Rahimi and Santosh Ramakrishna
Department of Computer Science, Southern Illinois University
US - Illinois - 62901, mail code 4511, Carbondale, USA
E-mail: rahimi@cs.siu.edu, srama@cs.siu.edu

The increasing complexity and difficulty of system administration has been long recognized. Studies indicate that because of the difficulty and complexity, the cost of administering systems is ten times the cost of the actual hardware. Here, we present ABSA; ABSA is an agent-based solution to automated system administration. ABSA architecture is introduced to minimize the cost of administering computers in multi platform networks and to provide a simple, consistent, expandable and integrated system administration tool. ABSA system supports important system administration features such as domain-wise administration, automated error handling and default system configuration besides others. (pp. 41-49)

In many intellectual or industrial fields, it is frequent that groups of actors have to collaborate in order to look for and to find solutions to new or old problems. The geographical distribution of the users imposes the use of a common software platform as transparent as possible. Having stated the hypothesis that the emergence of new solutions could arise by the co-location of potential solutions, we propose a lightweight peer to peer architecture based on mobile agents which implement a model of mobility called actual mobility (pp. 51-60)

Keywords: mobile agents, actual mobility, peer to peer application, intermediate design object.

Henry Hexmoor and Srinivas Battula
Computer Science and Computer Engineering, University of Arkansas,
US - AR 72701, Fayetteville, USA.
E-mail: Hexmoor@uark.edu, sbattul@uark.edu

In this paper we offer case studies of empowering agents with adjustment of cognitive notions of autonomy and trust that enable them to have a more socially adept interaction with a human supervisor. The application domain is control of unmanned aerial vehicles. Agents learn to change autonomies as they observe they learn the relationship between their assumed autonomy and performance. Agents also learn to change their reliance on human supervision as it takes different lengths of time. (pp. 61-67)

Keywords: Agents, Autonomy, Trust, Collaboration, Help

Marcin Paprzycki, Austin Gilbert, Andy Nauli, Minor Gordon, Steve Williams and Jimmy Wright
Computer Science Department
Oklahoma State University
US - OK 74106, Tulsa, USA
E-mail: marcin@cs.okstate.edu , austirg@cs.okstate.edu , nauli@cs.okstate.edu , minorg@cs.okstate.edu , stw@cs.okstate.edu , jimmyww@cs.okstate.edu

In this paper we discuss the problem of indexing information available on the Internet with the ultimate goal of delivering personalized content to users of an Internet-based travel support system. We introduce the form of index tokens that will be stored in the system and describe an agent-based subsystem designed to support the indexing function. Finally, we discuss the search agent that was developed to provide the system with index tokens and fueled experimentation with the proposed design. (pp. 69-78)

Keywords: content management, Internet, software agents, data indexing, ebXML registry/repository

Frederick Sheldon and Thomas Potok
Oak Ridge National Laboratory
Computational Sciences and Engineering
US - Tennessee 37831-6363, Oak Ridge, USA
Phone: 865-576-1339/ Fax: 865-241-6275
E-mail: SheldonFT@ornl.gov , PotokTE@ornl.gov

Krishna Kavi
Department of Computer Science
The University of North Texas
US - Texas 76203, Denton, USA
Phone: 940-565-2767 / Fax; 940-565-2799
E-mail: Kavi@cs.unt.edu

On the basis of three different agent-based development projects (one feasibility study, one prototype, one fully fielded), we assess the fitness of software (SW) agent-based systems (ABS) in various application settings: (1) distributed command and control (DCC) in fault-tolerant, safety-critical responsive decision networks, (2) agents discovering knowledge an open and changing environment, and (3) light weight distributed data management (DM) for analyzing massive scientific data sets. We characterize the fundamental commonalities and benefits of ABSs in light of our experiences in deploying the different applications. (pp. 78-89)

Keywords: Intelligent software agents, ontology, information fusion, collaborative decision support.

Rocchi Paolo
IBM
via Shangai 53
IT - Roma, Italy
E-mail:paolorocchi@it.ibm.com

This paper discusses the possibility of interpreting redundant information beyond the particular views emerging in specialist sectors. We introduce a theoretical framework that aims at unifying and calculating the main features of redundant information. This theoretical layout has been introduced in professional tuition. (pp. 91-94)

Keywords: Redundancy, control, information theory, reliability theory

Eva Jereb, Teja Toman
University of Maribor
Faculty of Organisational Sciences
Kidričeva cesta 55a
SI - 4000 Kranj, Slovenia
E-mail: eva.jereb@fov.uni-mb.si
WWW: http:// www.fov.uni-mb.si/eva

Nowadays the tempo of modern life is very fast which prevents parents from their regular contact with schools. We have developed an application, more precisely, an electronic grade (e-grade) book which enables the parents to monitor and control their child's education. By entering the username and password the parents would have an insight into their child's grades, inexcusable absence, test dates, cultural, sport days, natural science days, teachers' notes to the parents and so on. This paper shows an example of an e-grade book on the internet as well as the advantages and disadvantages a possible introduction of an e-grade in primary schools might bring about. Teachers' and parents' response to a possible introduction of an e-grade book is also presented. (pp. 95-101)

Keywords: internet, modelling, electronic grade system

N. Raja and R.K.Shyamasundar
School of Technology & Computer Science
Tata Institute of Fundamental Research
IN - Mumbai 400 005, INDIA
E-mail: raja@tifr.res.in, shyam@tifr.res.in

We explore the role of types in models of concurrent computation, particularly in the concrete setting of the asynchronous Π-calculus. The major theme of this work may be summarized by the slogan``Wherever you see structure, think of types'}. We propose type annotations not merely to channels, but also to the highly structured set of processes. The type system guarantees that well typed expressions cannot go wrong. Polymorphic process types formalize extant informal ideas regarding the channel passing and process passing approaches to process mobility. Further, subtyping relation between process types distinguishes between true concurrency and nondeterministic choice. (pp. 103-113)

Keywords: Type theory, concurrency, process calculi