Acquisition of Knowledge


ACKNOWLEDGE - 2576

Keywords knowledge-based systems, knowledge engineering, knowledge acquisition


Start Date: 01-JAN-89 / Duration: 39 months

[ contact / participants ]


Objectives and Approach

The aim of ACKNOWLEDGE is to improve the efficiency with which knowledge is acquired at all phases of knowledge-based system (KBS) development.

The objective is to construct a Knowledge Engineering Workbench (KEW) that will provide an integrated environment for knowledge elicitation and acquisition. Methods for integrating several knowledge acquisition techniques will be developed, and an environment that integrates a number of existing automated and non-automated techniques constructed.

The key constituents of this workbench are machine learning, knowledge elicitation and dialogue techniques. Learning mechanisms will support automatic deduction and organisation of new knowledge after new information has been provided or a particular event observed. This capability is necessary to guide the decisions and reactions of the environment. Elicitation methods will be implemented in the environment and will thus define the protocols of interaction with the system. Mixed-initiative dialogue techniques, based on the recognition of intentions and the visualisation of knowledge bases, will provide the foundations for integrating learning mechanisms and elicitation methods in the same environment. The project will also develop a comprehensive framework for guidance at all stages in KBS development.

Progress and Results

A first experimental prototype, integrating knowledge acquisition tools for diagnosis and classification applications, has been built and demonstrated, and following evaluation and experiments, a fully integrated version of the workbench has been designed and implemented. Feedback from evaluation, experiments and application developments has been incorporated in the final version of KEW.

The workbench includes various knowledge acquisition tools, knowledge storage facilities, knowledge transformation tools, advice and guidance for application developers and user interface libraries.

Exploitation

CAP Sogeti Innovation is demonstrating the capabilities of knowledge acquisition techniques to its customers using the initial tool-kit: it has already contributed in conjunction with KADS to win large development contracts (aeronautics, oil industry). The prototype is considered for internal use.

Telefonica will use the results in the creation and maintenance of internal knowledge base systems (ie EXPERPACPPN).

Computer Expert Systems intends to exploit the results for enhancing internal productivity and by creating domain-specific tools.

The Universiteit van Amsterdam will exploit the results through publications, PhD dissertations and courses for industry, as will Nottingham University, with special emphasis on the dissemination and exploitation of ACKNOWLEDGE products.

A spin-off company has approached the University of Nottingham to incorporate KEW in its commercial offer.

SINTEF will use the results in many industrial projects (help desk, environment monitoring, telecommunication, road maintenance, etc).

The consortium is also jointly discussing the opportunity to develop and commercialise a "Mini-KEW".


CONTACT POINT

Mr Cleo Jullien
CAP GEMINI INNOVATION
Chemin du Vieux Chêne
F - 38240 MEYLAN
tel: + 33/ 76-764747
fax: + 33/ 76-764748

Participants

CAP GEMINI INNOVATION - F - C
COMPUTAS EXPERT SYSTEMS A/S - N - P
MARCONI COMMAND & CONTROL
SYSTEMS LTD - UK - P
UNIVERSITEIT VAN AMSTERDAM - NL - P
UNIVERSITY OF NOTTINGHAM - UK - P
SINTEF - N - P
TELEFONICA INVESTIGACION
Y DESARROLLO - E - P
UNIVERSIDAD POLITECNICA DE MADRID - E - A


ST synopses home page ST acronym index ST number index
All synopses home page all acronyms index all numbers index

ACKNOWLEDGE - 2576, December 1993


please address enquiries to the ESPRIT Information Desk

html version of synopsis by Nick Cook