manstis:
I have looked at drools-chance, and my understanding is that this module is
closely tied to PMML file format, I was more looking for a generic score
models (additive/weighted/...) that can be defined using excel (and at some
point through a custom UI).
laune:
I agree that this is not a core "engine" functionality, and can be built on
similar lines as the decision tables module. My original question was not
phrased properly :(
Now that we have covered the basics of my question, can either one of you
point to any work/discussion in the drools community in this direction?
Also, if I have some thoughts on this subject (creating this framework), who
would be the person(s) to brainstorm? Who would provide me with a go/no-go
for working on this?
manstis wrote
Doh, forgot to mention the drools-change module supports PMML.
Helps if I join the dots up :)
On 5 June 2012 20:55, Michael Anstis <michael.anstis@> wrote:
> There's an experimental module called
>
https://github.com/droolsjbpm/drools-chance
>
> There are various sources on the internet about using PMML for Score
> Cards: e.g.
>
http://adapasupport.zementis.com/2010/02/scorecards-in-pmml-primer.html
>
> Might be worth looking at - I'm sure Davide Sotty would welcome feedback.
>
> With kind regards,
>
> Mike
>
>
> On 5 June 2012 18:40, Wolfgang Laun <wolfgang.laun@> wrote:
>
>> Thanks!
>>
>> This would not be implemented at "the Engine level"; it would be a
(more
>> or less) complex
>> framework similar to what there is now (in Drools) for decision tables.
>> But the evaluation for such
>> a scorecard is much less predictable than the one for a decision table,
>> which can be
>> defined in terms of Drools Rule Language conditions, constraints and
>> actions.
>>
>> If there is a more or less straightforward way of defining constraints
>> and actions from a scorecard,
>> it might not be so much different from the approach used for decision
>> tables. But is there a
>> commonly agreed-upon way of deriving the those from a scorecard? Is
>> there
>> a commonly
>> agreed-upon abstract structure for S.C. as there is one for
>> spreadsheets?
>>
>> -W
>>
>>
>>
>> On 5 June 2012 19:28, vinodkiran <vinodkiran@> wrote:
>>
>>> Score Models / Scorecards
>>>
>>> The Score Model / Scorecards is a common model used in scoring - a type
>>> of
>>> statistical approach, per Predictive Analytics - to allow a subjective
>>> score
>>> to be defined against some criteria for some decisioning purpose (eg
>>> credit
>>> scoring).
>>>
>>> A scorecard represents an assessment, opinion, or prediction about the
>>> customer or prospect. In other decision metaphors such as decision
>>> tables or
>>> decision trees, the values for rule conditions and actions are set by
>>> the
>>> user. In a scorecard, the properties and values for rule conditions and
>>> the
>>> rule action are generated using the values entered in the scorecard.
>>> The
>>> attribute scores are summed to return an overall score and the
>>> resulting
>>> score. This overall score is then available for use by other rulesets,
>>> functions, decision tables, and decision trees.
>>>
>>>
>>>
http://books.google.co.in/books?id=9AqRlUOaoVIC&pg=PA142&lpg=PA14...
>>>
>>>
>>> --
>>> View this message in context:
>>>
http://drools.46999.n3.nabble.com/scorecard-metaphor-in-drools-tp4017756p...
>>> Sent from the Drools: Developer (committer) mailing list mailing list
>>> archive at
Nabble.com.
>>> _______________________________________________
>>> rules-dev mailing list
>>> rules-dev@.jboss
>>>
https://lists.jboss.org/mailman/listinfo/rules-dev
>>>
>>
>>
>> _______________________________________________
>> rules-dev mailing list
>> rules-dev@.jboss
>>
https://lists.jboss.org/mailman/listinfo/rules-dev
>>
>>
>
_______________________________________________
rules-dev mailing list
rules-dev@.jboss
https://lists.jboss.org/mailman/listinfo/rules-dev
Sent from the Drools: Developer (committer) mailing list mailing list archive at