If models can interact then the 2nd approach has advantages, and the advantages grow the
tighter the interaction. If they're independent then there's no need for the
increased complexity of possible rule interaction, and the higher resource usage,
especially if you're just running one model at a time.
--- On Thu, 9/3/09, Kumar Mynam <Kumar.Mynam(a)imail.org> wrote:
From: Kumar Mynam <Kumar.Mynam(a)imail.org>
Subject: [rules-users] Different knowledge bases Vs One knowledge base with use of agent
filters..?
To: "rules-users(a)lists.jboss.org" <rules-users(a)lists.jboss.org>
Date: Thursday, September 3, 2009, 3:58 PM
Hello All,
We have different clinical models thru
which we send patient
data(HeartRate, MeanBP, etc...)
and alert clinicians with different
kind of alerts based on
output of model.
every clinical model has certain
activation rules, alert
rules. We are planning to use Drools for implementing
those rules.
a .drl file is implemented for each
model and rule
type(activation rules, alert rules)
We are thinking about two different
kinds of implementations.
1) Create a KnowledgeBase for each
model and rule type. Now,
when we want to evaluate data
thru model-1, we insert the facts into
that particular model
knowledge base statefulsession,
and fire the rules.
2) Create one big KnowledgeBase for
all the models and rule
types. Here, when we want to
evaluate data thru clinical model-1,
we get a session out of
KnowledgeBase, insert facts,
while firing rules we pass in agent
filters to fire only
model-1 rules.
Does 2nd approach have any advantages
over the 1st approach.
If so, what are those advantages,
or a link to that page would be
greatly appreciated.
Thank you.
kumar.
-----Inline Attachment Follows-----
_______________________________________________
rules-users mailing list
rules-users(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/rules-users