> The problem domain is in telecommunication. The problem is as the following:
> There are several databases scattering around the internet which capture
> the services used by subscribers. Since the subscribers can be moving around
> using different terminals (i.e. wireline telepohone, IP terminal, wireless
> telephone, etc), and using different services (i.e. telephone call,
> multimedia, etc), I would like to track the user during specific time.
Track or predict ?
> >From the data in databases, I could find out the probabilities and prior
> probabilities.
So I guess you want indeed to predict. " X is now using IP terminal,
what will he probably use after that ?"
If you really already have some probabilities, what is the problem ? It
is certainly better than fuzzy sets, for example. Remember that by
decreasing confidence you have
logical deduction, probabilistic induction, fuzzy induction, crystal ball
But maybe I don't understand exactly what you want to do ;-)
Maurice
-----------------------------------------------------------
Maurice CLERC http://perso.wanadoo.fr/maurice.clerc/
Kiam kien ni iru ?
"La fin est dans les moyens", "The aim is in the means", "Das Ziel ist
der Weg"
------------------------------------------------------------