Re: [open-nars] the number of senses processed by NARS in same time.
Pei Wang
(1) The control mechanism of the current version is still in
preliminary form, and needs a lot of work to handle a large knowledge
base. Before the logic part becomes mature, I have little time on
control.
(2) Even when the control part is finished as planned, the system may
still not be what you want --- After all, NARS is not designed to
achieve the best performance in a specific problem. For that kind of
situation, a special purpose system almost always works better than a
general purpose system.
Therefore, if you are only interested in using NARS for your project,
it may be better to only use its logic part (the inference rules), not
the control part. For a specific application, it is usually possible
to find an algorithm, which is more efficient than the
"non-algoristic" manner of NARS.
For the future of NARS, there is a plan to separate the logic part and
the control part more clearly, so other people can only use part of
the system for various purposes. For instance, if the Prolog version
of NAL (http://sites.google.com/site/narswang/implementations) is
further developed into a stand-along inference engine, it can be
easily integrated with various types of control methanism.
Pei
On Wed, Jul 28, 2010 at 7:01 PM, Max <max.t...@gmail.com> wrote:
> I guess the main problem with NARS integration in other project, is
> that we can not process more than 5 sentences at a time during
> reasonable period of time.
> So we have to process facts in procedural manner using other tool to
> construct the work flow.
> I mean that we can not put all the sentences that should check the
> genotype and genotype in NARS and just run "is it ok?" request.
>
> Is it possible that in near future there will be this kind of option.
> I guess the problem is that NARS infers using all possible
> combinations of the input sentences.
> I hope that there could be some mechanism that could guide the
> combination mechanism towards the request solution.
> If there is no such mechanism could you please pay some attention to
> this kind of issue please.
>
> Thank you in advance.
>
> --
> You received this message because you are subscribed to the Google Groups "open-nars" group.
> To post to this group, send email to open...@googlegroups.com.
> To unsubscribe from this group, send email to open-nars+...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/open-nars?hl=en.
>
>