KNIxMAS - Aspasia-Systems.De



		Knowledge Shared XPS - Based Research Network Using Multi-Agent Systems

		home I
		KNIXMAS Server I

		Esprit 977113
		INCO 977113


		Group of Automatic Control and Technical Cybernetics - atk (Prof. Dr. sc. techn. H. A. Nour Eldin)

		Dr.-Ing. H.Joachim Nern
		tel.: +49 171 476 2998
		nern -at- knixmas.de

Due to the increasing demand of globally distributed information within complex researching and developing there is a need for intelligent self adapting global research and development networks, especially under GroupWare aspects.
The main objective to design and establish a framework (shell) for an intelligent global research network consisting of connected local (national) working and research networks. This framework will consist of distributed co-operating expertsystems (Intelligent Base Systems) with decision support modules for research management and intelligent knowledge retrieval provided by applied multi-agent systems. For the prototype the local research networks and workingspaces will be realised as Intranets in 5 different European countries (Germany, Finland, Hungary, Poland, Bulgaria). The centre of each local network will be an intelligent base system conceived as expert system with co-operating and communication facilities via Internet/Intranet. The entire knowledge will be shared over the whole network and managed by Intelligent Knowledge Brokers.
The main characteristics of the research network are the following: The entire network is supplied with AI-components and methods. Modularity and flexibility will be guaranteed by structuring and designing the whole system in an object oriented way. Adaptiveness and flexible adjustment will be provided by reflecting and evaluating user response and acceptance data via complex feedback and influence loops. Developing this research network as a modular AI-system ensures a high degree of transferability and adjustment capabilities. Using dynamic knowledge and databases and digital repositories controlled by intelligent knowledge brokers guarantees inter- and exchange of specialised knowledge. For the intelligent knowledge retrieval global and local standard multi-agent systems will be installed on heterogeneous platforms using CORBA standards. The benefit for the users will be decision supported research management as well as intelligent knowledge retrieval embedded in collaborative workingspaces. In particular, the following achievements will be attuned to the European user:

1. Enhancing efficient global/local group networking,
2. Shortening integration time of new individual research terms,
3. Flexible allocation and supervision of global research teams,
4. Flexible long term planning and research task scheduling,
5. Maintainability, continuity, tracing and documentation of global/local research teams.

The flexible modular structure of the framework gives the possibility to apply the system to others than technical or scientific areas. It will be universally applicable for several areas in industry and economics. At first it is planned to process the knowledge of high-tech areas like control theory and communication techniques. In addition several testing partners are involved to test the system in other than scientific areas (customising of SPP/(R/3) software, enterprise modelling, Tele teaching etc.).

The participating project partners are:

Univ.- Prof. Dr. sc.techn. H.A. Nour Eldin as project co-ordinator, Institute of Automatic Control and Technical Cybernetics, University of Wuppertal, Germany, (Intelligent Base System (IBS) realised as co-operating expertsystem)
Prof. Dr. sci. J. D. Zaprianov as scientific co-ordinator, Institute of Control and Systems Research, Bulgarian Academy of Sciences, Bulgaria, (Decision Support Unit as research management system, Dynamic Knowledge Data Base)
Prof. Dr. hab. A. Dziech, Telecommunication Department, AGH ,University of Mining and Metallurgy , Krakow, Poland, (GroupWare for collaborative researching, Mobile Agent Architectures, CORBA-Standards
Prof. Dr. Laszlo Kovacs, MTA SZTAKI, Computer and Automation Research Institute of the Hungarian Academy of Science, Hungary, (Searching and Filtering in the WWW, Mobile Agent Architectures)
Prof. Dr. Matti Hämäläinen, Espoo-Vantaa Institute of Technology, Espoo, Finland, (Intelligent Knowledge Broker, market mechanisms, Document Manager)

1.2 Objectives and Results of the Project

The main objective of this project is to design and establish a framework (shell) for an intelligent global research network consisting of communicating local (national) working networks connected via Internet. The objective is to provide a comprehensive tool for collaborative working globally (international) and locally (national) in addition with decision supported researching and developing.
This framework consists of co-operating expertsystems (Intelligent Base System) with decision support modules for research management and intelligent information and knowledge acquisition based on multi agent systems. The application will be carried out as a national and international computer research network in the co-operating countries. The aim of this investigation and development is to create a framework-tool to increase and stabilise the research and technical development potential of eastern countries and moreover to provide a development framework, which is universal applicable in European areas due to its modular and transferable structure. The proposed system has flexible modular structure using well known system standards. For the present it is planned to implement particularly the knowledge of High-Tech relevant technologies, especially control theory, multivariable system theory and multidimensional signal processing. The contents of the flexible database will be transferable and interchangeable.
The following objectives and results of the project divided into specific measurable subobjectives are:

A Development of the framework of the Intelligent Base System(IBS)
Development of a network concept based on DESYS (atk)
Development of a concept for the supervision XPS based on existing XPS structures
Design and realisation of a rule base for generating and handling the ACCU, DSU and IKB
Design and realisation of an interface to the chosen MAEs
Concept and realisation of an user interface as a an administrator level

B Development of a general Decision Support Unit (DSU) as research management system
Analysis of the necessary properties of the rule base
Conception and realisation of a decision based research guidance module
Realisation of a module for representation of methodologies and research strategies

C Development of a Dynamic Knowledge Data Base (DKDB) and digital repositories
(BG) Development of a concept of appropriate knowledge representation
Design and realisation of units and shells for representation of: models, documents,
programme code, urls, graphics, experimental data, verification data etc.

D Design and realisation of an Agent Communication Co-ordination Unit within the IBS
(BG) Development of a concept of the ACCU as an interface between local and global MAEs
Realisation of the ACCU
Design and realisation of a module for transmitter functions and structures between global and local MAE
Realisation of adaptable blackboard structures and an agent dispatcher
Design and realisation of a mediator space for communication of local and global agent requests
Contractor mechanisms for different agent demands

E Realisation of an Intelligent Knowledge Broker (IKB) for handling the DKDB
Specifying market mechanisms as inference rules for sharing and exchanging information
Development an adequate concept for knowledge representation
Design of an document manager and applying meta -information schemes
Development of different template structures
Design of a module for database consistence checking

F MAEs at local side and collaborative working (groupware)
Installation of different MAEs (TCL, AGLETS, ARA etc ) ; testing, proof of applicability
Installation of MAEs (TCL, ARA; IABE etc) on heterogeneous platforms
Selection of a special MAE and CORBA oriented modification resp. standardisation
Design/synthesis of local feedback and adaptive mechanisms (feedback port)
Design/synthesis of local main working ports, controlled sharing of knowledge
Design/synthesis of local system service ports
Groupware approach for collaborative researching in local networks using MAEs
Providing standardised communication (CORBA) models of mobile agents for local networks

G User interfaces and adequate screen designs
Investigation visual relevant communication models
Investigation of integration of multi media components

H MA Installation of different MAEs (TCL, AGLETS, ARA etc ); testing, proof of applicability
Installation of MAEs (TCL, ARA; IABE etc) on heterogeneous platforms
Selection of a special MAE and CORBA oriented modification resp. standardisation
Design/synthesis of local feedback and adaptive mechanisms (feedback port)
Design/synthesis of local main working ports, controlled sharing of knowledge
Design/synthesis of local system service ports
Analysis of existing Search Agent program packages
Development of special searching methods using intelligent agents and subagents

1.3 Description of the Research Network

It is planned to develop an OnLine Research Network consisting of globally distributed Intelligent Base Systems IBS which are embedded in local research networks. The IBS are structured as expertsystems. The entire (international) research network is designed as a global network (via Internet) consisting of Local (national) Research and Working Networks LRWN (via Intranet and/or Internet). The communication, especially the knowledge retrieval (exchange) and requests (Intelligent Knowledge Acquisition) between the LRWNs is provided by applying Multi-Agent-Structures.
The centre of each LRWN is the special designed co-operating IBS with coordination capability consisting of a Local and a Global Multi-Agent-Environment LMAE/GMAE, an Intelligent Knowledge Broker IKB, Supervisor Expertsystem SXPS, Dynamical Knowledge- and Data Base DKDB, Decision-Support-Unit DSU, Agent-Communication-Coordination-Unit ACCU and an Administrator-Interface ADI. On the local side the IBS is connected to several local Work-Research-Spaces WRS established by the individual users (researchers). The WRS are designed similar to the IBS with respect to the individual requirements of the individual researcher (Individual Knowledge Data Base, Decision Support Unit with special problem oriented services, downloaded from the IBS) - a decision support service in the sense of a specialized research management. In the LRWNs the knowledge information exchange is either realized as direct Intranet connections and/or especially information retrieval agents with feedback capabilities (CORBA-standards) to fulfil the groupware aspect. The communication models of the different agents are oriented on CORBA-standards - allowing the use of different multi-agent-environments on heterogeneous platforms.
Thus the LRWNs are concepted as intelligent collaborative environments for the individual user. The comprehensive knowledge is shared over the globally distributed IBSs whereby each DKDB is specialised in substance (with regard to the contents).

1.3.1 IBS - Intelligent Base System

The main components of the IBS are:

* DKDB Dynamical Knowledge Data Base
* IKB Intelligent Knowledge Broker
* SXPS Supervisor Expertsystem
* DSU Decision Support Unit
* ACCU Agent Communication Coordination Unit
* LMAE Local Multi-Agent-Environment
* GMAE Global Multi-Agent-Environment

The DKDB contains the knowledge about a specific research area (it is planned for the first step: control theory and communication techniques, multivariable systems and multidimensional information processing) and it is structured in Docs, Math Models, Progr Code, Pics, URLs etc. The knowledge is prepared and treated with respect to an adequately chosen knowledge representation. The management (preparation, treatment, processing) of the DKDB is fulfilled by the IKB.
The main tasks of the IKB are: 1) representation of automatically (by global or local agents) and manually (by the administrator) acquired knowledge, 2) problem and goal oriented selection and filtering, 3) transformation of information based on given full structured templates, 4) management of rich text docs, 5) structuring and reduction of knowledge, 6) maintenance and access capabilities by the administrator using element profiler, 7) clustering of research goals and objectives, 8) insurance of knowledge consistence, 9) specifying market mechanisms for the information to be shared, 10) meta information schemes 11)controlling the input/output.

1.4 Benefit for the Users

As mentioned in the previous chapter the benefit for the users will be decision supported researching and development facilities as well as intelligent knowledge retrieval embedded in collaborative working spaces.
In particular, the following achievements will be attuned to the European user:
1. Enhancing efficient global/local group networking,
2. Shortening integration time of new individual research terms,
3. Flexible allocation and supervision of global research teams,
4. Flexible long term planning and research task scheduling,
5. Maintainability, continuity, tracing and documentation of global/local research teams.

At first it is planned to process the knowledge of high-tech areas like control theory and communication techniques. The flexible modular structure of the framework gives the possibility to apply the system to others than technical or scientific areas. It will be universally applicable for several areas in European industry and economics. This aspect will be considered due to contacts to European industry and economics for applying the system in European industry and economic areas. This will be shown in several test phases (see Exploitation chapter 4.4), where industry testing partners will test either the entire system or special components regarding other than scientific needs.