Nabble - xAIML

basic questions

2006-10-07T01:32:24Z

Dear all,

For a presentation on 1st december, I want to set up a chatbot in a
server, have a group of people train it and in the presentation
have the script talk toghether in the IRC with the lecturers and public.

I have a couple of questions:

1- would the group of people be able to 'train' the bot when the
programD is installed? (they are artists, and teachers, so it must
happen thought the brouwser)
2- I see that the IRC listener is only ready to chat with private
messages. Is there an upgrade so it can chat with the group?

I am very thankfull for any information,

claudia

_______________________________________________
xaiml mailing list
xaiml@...
http://aitools.org/mailman/listinfo/xaiml

Re: Multiple passes, infinite loops

2006-06-01T08:45:15Z

I wrote:

> just a global counter that makes sure that *no* category
> gets matched more than n times

And that's still too complicated. Just give me a way to configure an overall maximum of recursions, and I'll figure out whether I need 16, 128, whatever, for a given application.

Dirk

--
_______________________________________________

Search for businesses by name, location, or phone number. -Lycos Yellow Pages

http://r.lycos.com/r/yp_emailfooter/http://yellowpages.lycos.com/default.asp?SRC=lycos10

_______________________________________________
xaiml mailing list
xaiml@...
http://aitools.org/mailman/listinfo/xaiml

Re: Multiple passes, infinite loops

2006-06-01T05:03:23Z

Dirk Scheuring wrote:
> Noel Bush wrote:
>
>> This isn't a new idea -- basically corresponds to letting the
>> categories have named argument lists.
>
> Sorry, I didn't want to imply that this was my idea at all - it's a basic Lisp feature. I just wanted to show the users how to achieve the effect until the interpreter developers get around to implementing this feature formally, and show the interpreter devs that this might be one of the features worth considering.

No, no...sorry -- I was trying to stress that the alternate form that I
showed wasn't *my* idea. Didn't mean to seem to be suggesting that you
were implying that something that wasn't your idea was. :-)
_______________________________________________
xaiml mailing list
xaiml@...
http://aitools.org/mailman/listinfo/xaiml

Re: Multiple passes, infinite loops

2006-06-01T05:02:38Z

Noel Bush wrote:

> I guess the question for xAIML is: do you see
> some way that the matter of loop detection could become part of the
> language spec itself, rather than one of these "left up to the
> interpreter" matters? Are you saying that you would want to be able to
> change the loop counter on a per-category basis? What would that look like?

I think you're moving too fast :-). Like I said, users of the AAA set and similar sets built after the classic stimulus/response model don't need those loops, and probably shouldn't be encouraged to mess with them. When I started out, I used to write stuff like

<category>
<pattern>REDUCE _</pattern>
<template>
<srai>REDUCE SOMETHING</srai>
</template>
</category>

If I'd load this into an AIML interpreter and ran it, it would be nice if the interpreter would say something like, "Sorry, you've created an infinite loop which is caused be the category with the pattern 'REDUCE _'", and do it right after the first loop completion. This, I think, might be a useful default behavior, considering what the majority of people is likely to need.

Matching patterns more than once, IMO, only makes sense for the minority of AIML users who experiment with models different from s/r; in particular those of us who write functional programs in AIML. Kim Sullivan does this, I do this, and whoever found inspiration in the examples in the alicebot-style list archives might be doing this. Anybody here doing it? Anyway, at the mo, we should be a small user base. So any interpreter dev will probably consider whether delivering anything more than the bare necessities for those few people is worth their time.

With "bare necessities", I didn't mean "per-category-definable looping", but just a global counter that makes sure that *no* category gets matched more than n times, where n could be whatever a particular application needs (only the writer can determine how often is "more often than my code could possibly need"). The nth+1 recurrence should auto-match a pre-defined category that does *not* contain a srai in its template. This appears to be relativly simple to implement, generally useful, and safe.

Time-outs are another option; instead of fixing for a maximum depth, we'd have to estimate how long it should take to recurse n times. Are there advantages for you in implementing it that way?

Doing anything more will require the interpreter dev to consider the syntax and semantics of the functional languages that we implement using AIML. And this is where things get hairy, because my language is likely to be quite different from anybody else's language.

The general reason to do those loops is to be able to arbitrarily change the order in which the categories get matched, according to some theory or model that differs from the stimulus/response, head-over-tail, model for which the AIML matching algorithm was designed. The problem is that the language I create to do this depends in the content I want to encode, and therefore is different from writer to writer (and might even be different from project to project for one writer). So even though each one of us might be writing correctly terminating programs, most likely those are utterly incompatible, and maybe even incomparable, because they might be encoding very different models of, and assumptions about, natural language and its use by some target audience in the context of some content that is to be communicated. I'm afrais that we're in highly app-specific territory here.

Regrettfully, I don't have the slightest idea at the mo of how to remedy this unfortunate situation - how can we cope with this on the interpreter-development end, without making the devs think about the content we want to encode, and the various ideas about "grammar" and/or "matching prioritiy" that can be implemented using AIML? Anybody else got a hunch?

Dirk

Dirk

--
_______________________________________________

Search for businesses by name, location, or phone number. -Lycos Yellow Pages

http://r.lycos.com/r/yp_emailfooter/http://yellowpages.lycos.com/default.asp?SRC=lycos10

_______________________________________________
xaiml mailing list
xaiml@...
http://aitools.org/mailman/listinfo/xaiml

Re: Multiple passes, infinite loops

2006-06-01T02:40:40Z

Noel Bush wrote:

> This isn't a new idea -- basically corresponds to letting the
> categories have named argument lists.

Sorry, I didn't want to imply that this was my idea at all - it's a basic Lisp feature. I just wanted to show the users how to achieve the effect until the interpreter developers get around to implementing this feature formally, and show the interpreter devs that this might be one of the features worth considering.

>
> > As for the question of what an AIML interpreter developer can do
> > to detect the wide varity of infinite loops that I can generate
> > by effing up on tactics like this, in particular once I add
> > nested recursion - I'd rather they wouldn't bother. I just want
> > an interpreter with a simple loop counter that I can configure so
> > that it matches the same category at most m times, and then emits
> > a user-definable string. Loop depth should be set to "1" by
> > default, and the configuration code should be preceded by a
> > comment saying something like "Increase this value at your own
> > risk!"
>
> For some reason in Program D I removed the counting of loops, and just
> introduced a timeout. Since either seems so arbitrary and "magic
> numberish", I guess I figured that rather than encouraging people to try
> to guess how many times would be too many (a very tough problem), I'd
> just let the cut-off be configurable in terms of response time, which is
> going to be what really matters pragmatically. You can do a *lot* of
> recursions in a matter of a few milliseconds, but what really counts is
> whether you can "ever" come up with the "right" answer -- both "ever"
> and "right" being in scare quotes because "ever" almost always means "in
> a pretty short time" and "right" means "acceptable within a pretty short
> time"...or something like that.
>
> But I fully appreciate what you're saying about not trying to out-think
> you. As a general principle I think it is absolutely right -- *you* are
> the AIML wizard, not me. I guess the question for xAIML is: do you see
> some way that the matter of loop detection could become part of the
> language spec itself, rather than one of these "left up to the
> interpreter" matters? Are you saying that you would want to be able to
> change the loop counter on a per-category basis? What would that look like?
>
> > Please, as an interpreter developer, don't try to out-think me on
> > that regard - let me have the power I need, and don't disturb my
> > loops. As a tool for linguistic exploration, AIML becomes
> > worthless if you can't do "dangerous" things with it. Groping for
> > the cutting edge can't be perfectly safe, but you can always add
> > safety catches written in AIML, once you know, from enough
> > experience, how you can plunge into a particular abyss ;-)
> >
> > Dirk
> >
> _______________________________________________
> xaiml mailing list
> xaiml@...
> http://aitools.org/mailman/listinfo/xaiml

>

--
_______________________________________________

Search for businesses by name, location, or phone number. -Lycos Yellow Pages

http://r.lycos.com/r/yp_emailfooter/http://yellowpages.lycos.com/default.asp?SRC=lycos10

_______________________________________________
xaiml mailing list
xaiml@...
http://aitools.org/mailman/listinfo/xaiml

Re: Is this a feature request forum?

2006-05-31T22:43:00Z

charles chevallier wrote:
> Hi christmass oh.... sorry noel!

Hi Chuck -- I mean Charles,

All these sound like interesting ideas. I hope to make it clear that
xAIML itself will not contain any particular functionality, just will
provide a useful means for you to implement your extensions, me to
implement mine, and us to share (x)AIML and understand immediately which
parts our respective interpreters will and will not be able to handle.
You can add all the cool new features you want, and describe them in
xAIML, in a way that will also allow me to more easily try them out and
think about whether I want to support them in my interpreter.

I know that from some points of view this looks kind of insane -- there
are always so many pressures to standardize, to agree on a common set of
functionality. But in this particular world, what has become very clear
to me is that *nobody* knows what the real core set of functionality
ought to be...yet people have found standardization so valuable that
they rightly hesitate to deviate from the standard ("working draft"
though it may be). So this is an attempt to provide a way forward that
retains as many benefits of standardization as possible -- automating
the process of specification somewhat -- while at the same time opening
up the field to implement functionality to the same degree that any
programming language does the same (you build on top of class libraries,
etc.).

So we may wind up with a thousand xAIML dialects, yet "magically" they
can live harmoniously with one another, and xAIML authors are able to
navigate the field of possibilities with confidence and ease.

Something like that.

> It is me charlix the pinguin.
> I really love aiml, it is simple and already super powerful with the
> current set created by Richard Wallace.
> I definitly find great that you guys want to improve it.
>
> # compatibility with the standard
> I hope there could be a way to release substituting aiml categories for
> interpreters wich cannot handle new implementations like:
> <aiml>
> standard categories
> <xaiml_*>
> new specification * categories which would overwrite the previous ones
> and which would be ignored by other interpretaters
> </xaiml_*>
> </aiml>
>
> # implementation the pandorabot <learn> and <eval>
> Personnaly I don't think xAIML can be great for things like <sr2>
> because we are already able to use <srai><star index=1></srai>.
> But there are things that would be cool, like <learn><get
> name="aimldirectory"/>/update.aiml</learn> is not accepted with the
> standard. Moreover if you look systembadanswer.aiml in my files. I have
> to use sed and some operating system specific commands to add
> categories. (I know that for online and multi-user interpreters, the
> dynamic learning process is not as simple) But I think that there should
> be some ways to improve things, the pandorabot <learn> and <eval> are
> interesting.
>
> # implementation <set name=topic2> which comes before <pattern>
> Another thing is "_" I almost never use it because you cannot implement
> exceptions. <pattern>_ charlix</pattern><template><sr/></template>
> will be selected before <pattern>_ stupid
> charlix</pattern><template><sr/></template>
> So I thought about # which would match backwards from the end to the
> begining like <pattern># charlix</pattern><template><sr/></template>
> will be selected after <pattern># stupid
> charlix</pattern><template><sr/></template>
> Or I thought about a tag <nrettap> pattern backward.
> At last I thought the most compatible way with the aiml standard is to
> add in a preprocessed topic <pattern>*
> charlix</pattern><template><sr/></template>
> will be selected before <pattern>* stupid
> charlix</pattern><template><sr/></template>
> Unfortunatly <topic> and <that> will be selected after <pattern>.
> So why not a <topic2> which would be selected before <pattern> or even a
> <topic3> which would be exclusiv then topic2 can be implemented by
> adding <set name=topic3></set><sr/>
>
> Pardon my french, I hope my english is understandable
>
> # reversedaiml
> By the way you can have a look at my new experimental release
> reversedaiml.aiml. it returns for a simple sentence the possible
> questions inside aiml tags. so that you can transform any plain english
> text into learnable aiml. So far only sentences with irregular verbs
> will be correctly processed. As charlix, it is GPLed so if you maintain
> any open source interpreters, you are more than wellcome to embed
> charlix or reversedaiml. (I'd like to release a windows demo of charlix.
> <system>start iexplore</system> works but <system>start
> anyprogram.exe</system> hangs. so if you have a solution please email
> me. Finally, Pandorabot cannot handle enough recursions for
> reversedaiml, so if you have a hosting solution please go on, publish it.)
>
>
> Thanks guys to create our supertoys' tools
>
> Charlix
>
> ------------------------------------------------------------------------
> *Yahoo! Mail réinvente le mail !* Découvrez le nouveau Yahoo! Mail
> <http://fr.rd.yahoo.com/evt=40577/*http://fr.promotions.yahoo.com/mail/nouveaumail.html
> > et son interface révolutionnaire.
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> xaiml mailing list
> xaiml@...
> http://aitools.org/mailman/listinfo/xaiml

_______________________________________________
xaiml mailing list
xaiml@...
http://aitools.org/mailman/listinfo/xaiml

Re: Multiple passes, infinite loops

2006-05-31T22:35:50Z

Dirk Scheuring wrote:
> This is easy to generate; no extension of AIML is necessary to do in-topic multi-pass, or multi-topic multi-pass, or even branched multi-topic multi-pass transformations. It's a powerful linguistic tool; I'm a long ways from having explored all of its possibilities, but I know they are there.

Of course the implementors of various xAIML dialects may have many
different interests, but for those such as I who are interested in
experimenting with extensions that operate at a fairly "core" level, I
would approach the issue you're describing by trying to offer some
semantics that would enable you to do exactly what you're illustrating
here but with more explicitness about what's going on. So, for
instance, one thing that seems highly desirable is to separate
attributes that have to be "encoded" in match strings and patterns, into
actual labeled attributes. One might wish to match not just string
patterns, but attribute patterns. So your example might become:

<category>
<pattern>* WORD *</pattern>
<template>
<srai>
<string id="message"><star/> WORD <star index="2"/></string>
<string id="sense">1</string>
</srai>
</template>
</category>

<category>
<pattern>
<string id="message">* WORD *</string>
<string id="sense">1</string>
</pattern>
<template>
<srai>
<string id="message"><star/> WORD <star index="2"/></string>
<string id="sense">2</string>
</srai>
</template>
</category>

....

<category>
<pattern>
<string id="message">* WORD *</string>
<string id="sense">*</string>
</pattern>
<template>
<srai>
<string id="message"><star index="2"/></string>
</srai>
</template>
</category>

Or something along those lines. This isn't a new idea -- basically
corresponds to letting the categories have named argument lists.

> As for the question of what an AIML interpreter developer can do to detect the wide varity of infinite loops that I can generate by effing up on tactics like this, in particular once I add nested recursion - I'd rather they wouldn't bother. I just want an interpreter with a simple loop counter that I can configure so that it matches the same category at most m times, and then emits a user-definable string. Loop depth should be set to "1" by default, and the configuration code should be preceded by a comment saying something like "Increase this value at your own risk!"

For some reason in Program D I removed the counting of loops, and just
introduced a timeout. Since either seems so arbitrary and "magic
numberish", I guess I figured that rather than encouraging people to try
to guess how many times would be too many (a very tough problem), I'd
just let the cut-off be configurable in terms of response time, which is
going to be what really matters pragmatically. You can do a *lot* of
recursions in a matter of a few milliseconds, but what really counts is
whether you can "ever" come up with the "right" answer -- both "ever"
and "right" being in scare quotes because "ever" almost always means "in
a pretty short time" and "right" means "acceptable within a pretty short
time"...or something like that.

But I fully appreciate what you're saying about not trying to out-think
you. As a general principle I think it is absolutely right -- *you* are
the AIML wizard, not me. I guess the question for xAIML is: do you see
some way that the matter of loop detection could become part of the
language spec itself, rather than one of these "left up to the
interpreter" matters? Are you saying that you would want to be able to
change the loop counter on a per-category basis? What would that look like?

> Please, as an interpreter developer, don't try to out-think me on that regard - let me have the power I need, and don't disturb my loops. As a tool for linguistic exploration, AIML becomes worthless if you can't do "dangerous" things with it. Groping for the cutting edge can't be perfectly safe, but you can always add safety catches written in AIML, once you know, from enough experience, how you can plunge into a particular abyss ;-)
>
> Dirk
>
_______________________________________________
xaiml mailing list
xaiml@...
http://aitools.org/mailman/listinfo/xaiml

Re: Is this a feature request forum?

2006-05-30T05:39:05Z

--- charles chevallier <chcdsl@...> escreveu:

> I hope there could be a way to release substituting
> aiml categories for interpreters wich cannot handle
> new implementations like:

Well, xAIML is pretty much all about enabling
interpreters to "fail gracefully" when finding
extensions they can't handle. Also see AIML's
"forward-compatible processing" feature:

http://www.alicebot.org/TR/2005/WD-aiml/#section-forward-compatible-processing

--
Ja mata ne.
Helio Perroni Filho

_______________________________________________________
Novo Yahoo! Messenger com voz: Instale agora e faça ligações de graça.
http://br.messenger.yahoo.com/
_______________________________________________
xaiml mailing list
xaiml@...
http://aitools.org/mailman/listinfo/xaiml

P.S.: charlix website

2006-05-29T15:59:48Z

sorry I forgot

if you don't know me charlix the penguin, look here:

http://charlix.sourceforge.net/

thank you

Charlix

Yahoo! Mail réinvente le mail ! Découvrez le nouveau Yahoo! Mail et son interface révolutionnaire.
_______________________________________________
xaiml mailing list
xaiml@...
http://aitools.org/mailman/listinfo/xaiml

Is this a feature request forum?

2006-05-29T15:54:34Z

####################################
SUMARY OF THIS MAIL
# compatibility with the standard
# implementation the pandorabot <learn> and <eval>
# implementation <set name=topic2> which comes before <pattern>
# reversedaiml
######################################

Hi christmass oh.... sorry noel!

It is me charlix the pinguin.
I really love aiml, it is simple and already super powerful with the current set created by Richard Wallace.
I definitly find great that you guys want to improve it.

# compatibility with the standard
I hope there could be a way to release substituting aiml categories for interpreters wich cannot handle new implementations like:
<aiml>
standard categories
<xaiml_*>
new specification * categories which would overwrite the previous ones and which would be ignored by other interpretaters
</xaiml_*>
</aiml>

# implementation the pandorabot <learn> and <eval>
Personnaly I don't think xAIML can be great for things like <sr2> because we are already able to use <srai><star index=1></srai>.
But there are things that would be cool, like <learn><get name="aimldirectory"/>/update.aiml</learn> is not accepted with the standard. Moreover if you look systembadanswer.aiml in my files. I have to use sed and some operating system specific commands to add categories. (I know that for online and multi-user interpreters, the dynamic learning process is not as simple) But I think that there should be some ways to improve things, the pandorabot <learn> and <eval> are interesting.

# implementation <set name=topic2> which comes before <pattern>
Another thing is "_" I almost never use it because you cannot implement exceptions. <pattern>_ charlix</pattern><template><sr/></template>
will be selected before <pattern>_ stupid charlix</pattern><template><sr/></template>
So I thought about # which would match backwards from the end to the begining like <pattern># charlix</pattern><template><sr/></template>
will be selected after <pattern># stupid charlix</pattern><template><sr/></template>
Or I thought about a tag <nrettap> pattern backward.
At last I thought the most compatible way with the aiml standard is to add in a preprocessed topic <pattern>* charlix</pattern><template><sr/></template>
will be selected before <pattern>* stupid charlix</pattern><template><sr/></template>
Unfortunatly <topic> and <that> will be selected after <pattern>.
So why not a <topic2> which would be selected before <pattern> or even a <topic3> which would be exclusiv then topic2 can be implemented by adding <set name=topic3></set><sr/>

Pardon my french, I hope my english is understandable

# reversedaiml
By the way you can have a look at my new experimental release reversedaiml.aiml. it returns for a simple sentence the possible questions inside aiml tags. so that you can transform any plain english text into learnable aiml. So far only sentences with irregular verbs will be correctly processed. As charlix, it is GPLed so if you maintain any open source interpreters, you are more than wellcome to embed charlix or reversedaiml. (I'd like to release a windows demo of charlix. <system>start iexplore</system> works but <system>start anyprogram.exe</system> hangs. so if you have a solution please email me. Finally, Pandorabot cannot handle enough recursions for reversedaiml, so if you have a hosting solution please go on, publish it.)

Thanks guys to create our supertoys' tools

Charlix

Re: Welcome

2006-05-29T12:55:43Z

--- Noel Bush <noel@...> escreveu:

> It would be ideal, from my point of view, if those
> of us who are implementors of AIML interpreters
> would take a crack at prototyping xAIML-like
> functionality, so we and others could play around
> with the ideas and start seeing what works. I
> realize everybody's time is limited, as is mine,
> but in individual feedback I've gotten a pretty good
> sense that there will be several people up for
> trying this out.

Hello Noel,

As you pointed out, everybody's time is limited, hence
the delay in this response. ^_^' I agree that start
building xAIML handlers is the best way to see whether
and where there are problems with the concept, and to
figure out the best way to deal with specific issues.
Currently my checklist for ChatterBean is as follows:

1- Rework the input processing engine;
2- Implement the last remaining features in the AIML
standard;
3- Complete the JavaBean interface;
4- Start playing with xAIML.

Being a fairly new addition to the set of intended
features, xAIML doesn't score high in my to-do list,
but that doesn't mean I don't regard it as important:
it's just that I need to get the basic AIML right
before giving much more thought to extensions.
Meanwhile, I hope someone with a more mature AIML
engine can get to adding xAIML features to their
system, and share their experience with us.

--
Ja mata ne.
Helio Perroni Filho

_______________________________________________________
Abra sua conta no Yahoo! Mail: 1GB de espaço, alertas de e-mail no celular e anti-spam realmente eficaz.
http://mail.yahoo.com.br/
_______________________________________________
xaiml mailing list
xaiml@...
http://aitools.org/mailman/listinfo/xaiml

Multiple passes, infinite loops

2006-05-29T08:08:34Z

--- In Robitron@..., Noel Bush wrote:

> What's really needed, assuming one is coming from an AIML world,
> is to extend functionality such that you take multiple passes at
> an input, and break it up into individual utterances which are to
> be discarded or handled as appropriate, based not on the flimsy
> evidence of punctuation, but on plain old "content" -- same as
> AIML works now for individual sentences.

I've done some more thinking on this issue, and I guess it's more appropriate to discuss this here than on Robitron. First of all, there exists a way to do /safe/ multiple passes over an input while staying in one <topic>, by using what I call "self-modding words". For instance, what I can do to iterate over several possible interpretations, or "senses", of a word is to prefix the words I want to involve in the multi-pass transform with numbers, and use the numbered versions of the word in the various patterns that include that word in its different "senses".

<category>
<pattern>* WORD *</pattern>
<template>
<srai>
<star index="1"/> 1WORD <star index="2"/>
</srai>
</template>
</category>

<category>
<pattern>* 1WORD *</pattern>
<template>
<srai>
<star index="1"/> 2WORD <star index="2"/>
</srai>
</template>
</category>

<category>
<pattern>* 2WORD *</pattern>
<template>
<srai>
<star index="1"/> 3WORD <star index="2"/>
</srai>
</template>
</category>



<category>
<pattern>* NWORD *</pattern>
<template>
<srai>
<star index="2"/>
</srai>
</template>
</category>

This is easy to generate; no extension of AIML is necessary to do in-topic multi-pass, or multi-topic multi-pass, or even branched multi-topic multi-pass transformations. It's a powerful linguistic tool; I'm a long ways from having explored all of its possibilities, but I know they are there.

As for the question of what an AIML interpreter developer can do to detect the wide varity of infinite loops that I can generate by effing up on tactics like this, in particular once I add nested recursion - I'd rather they wouldn't bother. I just want an interpreter with a simple loop counter that I can configure so that it matches the same category at most m times, and then emits a user-definable string. Loop depth should be set to "1" by default, and the configuration code should be preceded by a comment saying something like "Increase this value at your own risk!"

Please, as an interpreter developer, don't try to out-think me on that regard - let me have the power I need, and don't disturb my loops. As a tool for linguistic exploration, AIML becomes worthless if you can't do "dangerous" things with it. Groping for the cutting edge can't be perfectly safe, but you can always add safety catches written in AIML, once you know, from enough experience, how you can plunge into a particular abyss ;-)

Dirk

--
_______________________________________________

Search for businesses by name, location, or phone number. -Lycos Yellow Pages

http://r.lycos.com/r/yp_emailfooter/http://yellowpages.lycos.com/default.asp?SRC=lycos10

_______________________________________________
xaiml mailing list
xaiml@...
http://aitools.org/mailman/listinfo/xaiml