|
Network
of Excellence
|
CARiMan
|
The
Internet Portal for Computer Aided Risk Management
|
KM:
Knowledge
Management and Data Base System
Responsible:
MIRA
Assistance:
CEFIN, DET,
HP,
all
partners
Objectives
and
system description
There are several types of information that the CARiManS will keep and
make available to the risk managers. The system has to understand the semantic change
in each particular monitored system, therefore, it has to be able to perform
semantic understanding of the change in signals, and has to be able to learn and
adapt to each monitored system and through time. In addition, the system is based on knowledge incorporated
from risk managers, i.e. on knowledge acquisition. Thus, there will be a
collection of heterogeneous types of information and formats that need to be
made accessible through intuitive interfaces.
This will include information in the traditional sense as well as
knowledge and knowledge records that can be considered as meta-information in
this case.
The thematic of this scenario requires a seamless
navigation capability by the user irrespective of which formats information is
in of what type of information is represented, such as raw data, interpreted
data and knowledge records. The thematic, when viewed from an information and
knowledge management perspective is closely related to that of the semantic web
where information irrespective of its original form or provenance can be made
navigable and retrievable according to the preferences of the information
consumer rather than the original provider.
Frequently observed problems are the information
redundancy and patchiness in the primary databases as well as heterogeneity of
data formatting. These facts presume
application of methods for linguistically determined and processed knowledge,
which in fact is a basis for implementation of the intelligent (soft computing)
methods and particularly fuzzy, rough and neural network modeling algorithms. By
means of these algorithms the subjective information received from the risk
managers will be implemented in the Data Base and Knowledge Management System.
As a base of such a system the Metagen Data Base can be
used. Moreover in order to obtain good communication risk manager - CARiManS for
acquisition of the knowledge of experts we propose investigation and
implementation of techniques based on modern web-based machine-to-machine XML
technologies. Here the web-based XML technologies will be exploited as well as
soft computing methods for data representation and motion control. The aim is
the risk managers to be able to observe distantly the operation done in other
place in order perform direct control on the management procedure, or to advise
the team doing the operation.
The implementation of the database will be conducted in
three steps. The first step will involve the definition of available data
sources and assigning the user’s access to relevant data. In this phase
activities will address the design of data mining and information integration
algorithms. The second step of implementation process will include the analysis
of the users’ needs in the area of access technologies. Activities undertaken
in the second phase will include also the definition of the contents preparation
rules to assure its suitability for browsing and viewing using different
equipment. Multiplication of possibilities will quickly cause growth of space
required for storing the same content prepared in different manner. Therefore,
the third phase will aim at preparation of universal risk management content,
which could be recompressed and reorganised on-the-fly in order to achieve data
being suitable for current user requirements (access network solution, device
type as well as personal preferences).