3:38 p.m.
Hi for all and sorry for any duplication.
My team are starting a project where we will develop a tool to visualize
metrics of a production rule base and show them in a 3D form.
Through this tool we want measure quality and conformity with good
practices to help in detection of anomalies and ensure a good design of a
knowledge base.
Other tools and studies already exists to visualize code and calculate
metrics from java code. Examples are CodeCity (
http://www.inf.usi.ch/phd/wettel/codecity.html<http://www.linkedin.com...),
where java code is visualized using a city metaphor and Sonar where metrics
are extracted from source code to help in code quality assurance (
http://nemo.sonarsource.org/dashboard/index/327690<http://www.linkedin...
).
So, my questions is:
1 - How much relevant you think this idea is?
2 - You know about other research in this area?
Thanks in advance.
--
----------------------------------------------------------------------------------------------------------------
Jackson Cunha Cassimiro (CereB)
Bacharel em Ciencia da Computação - UFPI
MSN: jackson.cereb(a)gmail.com
Telefone Móvel +55 86 9928 1251
Analista de Sistemas - Infoway - http://www.infoway-pi.com.br
Missão Infoway - "Influenciar a Gestão de Sistemas de Saúde através de
e-health"
("A vida é um combate que os fracos abate, aos bravos, aos fortes só pode
exaltar" - Canção do Tamoio, Gonçalves Dias)
----------------------------------------------------------------------------------------------------------------
7:31 p.m.
There are metrics that can be applied to Java code (or the code of similar
languages). Some of these metrics might be applied to the code of a rule's
right hand side, although it should be noted that RHS code should not use
conditional statements (if, switch, for) as this would undermine the
general production rule principle of "LHS selects - RHS executes".
Stringent coding rules would simply rule out if and switch for RHS code.
LHS code is written in an extended expression language. It's difficult to
judge RHS code based on the frequency of CEs, on the quality of CEs and the
number of constraints. There is one problem category I know that can be
solved using a single rule with a fairly complex CE combination.
Nesting "from" or "accumulate" results (as might be expected) in
complex
and not too efficient rules, but that's to be expected, and, occasionally,
it can't be helped.
What would your metrics be based on? Showing values in 3D may or may not be
a boon, depending on what the dimensions stand for.
Typically, computing metrics results in values that alert you to expect
more errors or higher meintencance effort due to complexity. Whether it's
possible to express the same logic with rules with smaller "danger" metrics
remains to be seen; sometimes it's possible by redesigning, i.e., what is a
fact, how facts are linked to each other, etc.
-W
On 7 May 2012 15:38, Jackson Cunha <jackson.cereb(a)gmail.com> wrote:
Hi for all and sorry for any duplication.
My team are starting a project where we will develop a tool to visualize
metrics of a production rule base and show them in a 3D form.
Through this tool we want measure quality and conformity with good
practices to help in detection of anomalies and ensure a good design of a
knowledge base.
Other tools and studies already exists to visualize code and calculate
metrics from java code. Examples are CodeCity (
http://www.inf.usi.ch/phd/wettel/codecity.html<http://www.linkedin.com...),
where java code is visualized using a city metaphor and Sonar where metrics
are extracted from source code to help in code quality assurance (
http://nemo.sonarsource.org/dashboard/index/327690<http://www.linkedin...
).
So, my questions is:
1 - How much relevant you think this idea is?
2 - You know about other research in this area?
Thanks in advance.
--
----------------------------------------------------------------------------------------------------------------
Jackson Cunha Cassimiro (CereB)
Bacharel em Ciencia da Computação - UFPI
MSN: jackson.cereb(a)gmail.com
Telefone Móvel +55 86 9928 1251
Analista de Sistemas - Infoway - http://www.infoway-pi.com.br
Missão Infoway - "Influenciar a Gestão de Sistemas de Saúde através de
e-health"
("A vida é um combate que os fracos abate, aos bravos, aos fortes só pode
exaltar" - Canção do Tamoio, Gonçalves Dias)
----------------------------------------------------------------------------------------------------------------
_______________________________________________
rules-users mailing list
rules-users(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/rules-users
11:23 p.m.
Thanks Wolfgang!
*
*
I already have some papers to read about complexity of production rules and
soon I will post here something about this.
But, answering my question: do you think this idea is relevant?
----------------------------------------------------------------------------------------------------------------
Jackson Cunha Cassimiro (CereB)
Bacharel em Ciencia da Computação - UFPI
MSN: jackson.cereb(a)gmail.com
Telefone Móvel +55 86 9928 1251
Analista de Sistemas - Infoway - http://www.infoway-pi.com.br
Missão Infoway - "Influenciar a Gestão de Sistemas de Saúde através de
e-health"
("A vida é um combate que os fracos abate, aos bravos, aos fortes só pode
exaltar" - Canção do Tamoio, Gonçalves Dias)
----------------------------------------------------------------------------------------------------------------
7:38 a.m.
I can't answer this w.r.t. the "3D" unless I know what the 3
dimensions stand for, can I?
Some questions that I can think of offhand that have a bearing on the
relevance of the idea of a production rule metric:
Can a metric capture the complexity resulting from the interaction of
multiple rules?
Can a metric reflect the trickiness of window:x in combination with
parameterized temporal operators?
Will various rule attributes (such as agenda-group, salience,...)
affect the metric?
Is the reevaluation effect of delete/modify/insert taken into account?
Cheers
Wolfgang
2012/5/7 Jackson Cunha <jackson.cereb(a)gmail.com>
Thanks Wolfgang!
I already have some papers to read about complexity of production rules and soon I will
post here something about this.
But, answering my question: do you think this idea is relevant?
----------------------------------------------------------------------------------------------------------------
Jackson Cunha Cassimiro (CereB)
Bacharel em Ciencia da Computação - UFPI
MSN: jackson.cereb(a)gmail.com
Telefone Móvel +55 86 9928 1251
Analista de Sistemas - Infoway - http://www.infoway-pi.com.br
Missão Infoway - "Influenciar a Gestão de Sistemas de Saúde através de
e-health"
("A vida é um combate que os fracos abate, aos bravos, aos fortes só pode
exaltar" - Canção do Tamoio, Gonçalves Dias)
----------------------------------------------------------------------------------------------------------------
_______________________________________________
rules-users mailing list
rules-users(a)lists.jboss.org
https://lists.jboss.org/mailman/listinfo/rules-users
4432
days inactive
4433
days old
3 comments
2 participants
participants (2)
-
Jackson Cunha -
Wolfgang Laun