Hi Anindya and others,
>
> I want to know what "KNOWLEDGE-BASED SOFTWARE SYSTEM FOR GEOLOGICAL
> STRUCTURE ANALYSIS" mean. What sort of work can be visualised and worked
> upon ?
I'm not that much a cyber guy of my own, but I'll try to convey what I
know.
The systems you point at possess learning algorithms that can be
instructed to gather information from data they process, ie pattern
recognition, target identification, gray-scale image analysis. If you
input, say, a seismic profile, then your interpretation (or a likely one,
even with errors), the system will aim at associating data patterns with
interpretation patterns, iterating the process for a given number of times
and on a given number of datasets. Apprenticeship of the algorithms builds
up a reference library for a given project, that therefore constitutes
'knowledge'. Next time the system is confronted with data of resemblable
structure and pattern and significance (on the ground of chosen
parameters), it will provide an interpretation based on what the system
already knows about, for the example above, fault pattern recognition. The
advantage is that sytems learns on the go, and embodies a substantial
amount of information for the following case.
These system don't come from the geological interpretation world, rather
from missile and target detection (for image processing) and for signal
processing purposes in general. The application to geologically-sound
patterns is rather recent, but it can be viable for very large data
arrays.. I suppose it is used in seismic processing of huge datasets,
whereas application to visual discrimination of, say, fault architecture
is still relatively anew, although studies exist. I remember a paper by
people in Abu Dhabi (if I am nto wrong..) with respect to reservoir
modelling, but I don't have it handy.. If I come up with further memories,
I'll write back.
Hope this can help a little bit, best of luck,
Umberto
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|