|
|
Senior member
Total posts: 974
Joined: Oct 21, 2009
|
|
|
|
|
|
Posted: Mar 10, 2011 |
[ # 1 ]
|
|
Administrator
Total posts: 3111
Joined: Jun 14, 2010
|
While Morti uses AIML architecture, I don’t consider him to be a “clone”. After all, his script is slowly expanding beyond the confines of AIML, and his next incarnation will likely be using some form of NLP for his input analysis before too much longer. I’m also strongly considering using some form of Bruce Wilcox’s ChatScript (IF I can port the code to PHP) to handle not only the input, but also to truly “generate” the output, rather than match an output to a detected input. I’ve got some thoughts on the subject, but I’ll talk about them in a new thread here in a few days. I’m still working on a couple of big projects, and they come first.
|
|
|
|
|
Posted: Mar 10, 2011 |
[ # 2 ]
|
|
Senior member
Total posts: 623
Joined: Aug 24, 2010
|
I think it’s a little rough having a category labeled “AIML clone”. Especially since using the AIML markup language is very different from making a copy of a bot and calling it your own. And Alice/AIML-based bots aren’t the only ones that gets cloned and re-branded.
|
|
|
|
|
Posted: Mar 10, 2011 |
[ # 3 ]
|
|
Senior member
Total posts: 697
Joined: Aug 5, 2010
|
You also have those inf (or how are they called) things these days
update: http://www.inf.net/
|
|
|
|
|
Posted: Mar 10, 2011 |
[ # 4 ]
|
|
Senior member
Total posts: 494
Joined: Jan 27, 2011
|
Great poll Victor, if only it acknowledges the fact that a project can be in a certain stage
I voted the second option of course.
|
|
|
|
|
Posted: Mar 10, 2011 |
[ # 5 ]
|
|
Administrator
Total posts: 2048
Joined: Jun 25, 2010
|
I too use AIML but don’t class my seven years worth of modifications as a clone.
|
|
|
|
|
Posted: Mar 10, 2011 |
[ # 6 ]
|
|
Guru
Total posts: 1297
Joined: Nov 3, 2009
|
We need a category for alicebot clones less the AIML:
http://79b.org/cgi-bin/RiveScriptLog.cgi
Suggestion: Alessbot
|
|
|
|
|
Posted: Mar 10, 2011 |
[ # 7 ]
|
|
Senior member
Total posts: 974
Joined: Oct 21, 2009
|
Ahhh…. good points Steve and CR. I should have expanded that. 8PLA, yes, we could have another pool for the specifics of AIML related bots.
Dave, that’s interesting that you are expanding Moriti like that.
Hans - yes, I put that option in there because I knew that is where you currently are in your project.
|
|
|
|
|
Posted: Mar 10, 2011 |
[ # 8 ]
|
|
Senior member
Total posts: 494
Joined: Jan 27, 2011
|
Victor Shulist - Mar 10, 2011: Hans - yes, I put that option in there because I knew that is where you currently are in your project.
I figured as much
|
|
|
|
|
Posted: Mar 10, 2011 |
[ # 9 ]
|
|
Administrator
Total posts: 3111
Joined: Jun 14, 2010
|
I’ve always felt (and I think that many others agree) that AIML is a great “Beginner’s Platform” for learning about chatbots and Artificial Intelligence. Where I seem to differ from the others is that, when a certain point is reached, many have simply “dropped” AIML, in preference for other platforms, or have started working on new platforms of their own. I, however, plan on expanding on AIML’s functionality as I learn the required skills, and morph Morti’s script over time into something more than what it is. At some point, when I create or discover a better way to generate output, I’ll likely phase the AIML portion out, but I may (or may not) keep some small amount of it’s functionality. For me, the most exciting part of the journey is the journey itself.
|
|
|
|
|
Posted: Mar 10, 2011 |
[ # 10 ]
|
|
Senior member
Total posts: 974
Joined: Oct 21, 2009
|
Dave, Agreed.
It is even possible, with someone with a novel approach, to add to AIML to give it even greater capabilities. My own opinion on AI, is ‘bottom up’ philosophy.
|
|
|
|
|
Posted: Mar 10, 2011 |
[ # 11 ]
|
|
Administrator
Total posts: 3111
Joined: Jun 14, 2010
|
I’ve heard a lot of folks promoting “top down”, and “bottom up” approaches to AI, and that’s great! I think the Vulcan acronym (translated into English, of course) I.D.I.C. Says it best, “Infinite Diversity in Infinite Combinations”. Me? I’m taking more of an “inside out and backwards” approach.
|
|
|
|
|
Posted: Mar 10, 2011 |
[ # 12 ]
|
|
Senior member
Total posts: 974
Joined: Oct 21, 2009
|
Dave, inside-out, thanks goooood ! In a way mine is a bit of an inside out approach as well.
|
|
|
|
|
Posted: Mar 10, 2011 |
[ # 13 ]
|
|
Senior member
Total posts: 494
Joined: Jan 27, 2011
|
Dave Morton - Mar 10, 2011: Where I seem to differ from the others is that, when a certain point is reached, many have simply “dropped” AIML, in preference for other platforms, or have started working on new platforms of their own.
I think AIML is a pretty solid foundation to build much more involved models. The experiments I did with implementing the PAD-model in AIML showed me that it is totally feasible to implement emotions in a AIML-bot. The only drawback to me is that XML is a rather bloated way to describe things, it’s like INI-files with loads of XHTML thrown in to confuse things.
I just recently found a pre-processor that let you write AIML without all the XML-tags and then generate a compliant AIML-file from it.
|
|
|
|
|
Posted: Mar 10, 2011 |
[ # 14 ]
|
|
Guru
Total posts: 1297
Joined: Nov 3, 2009
|
This is a good candidate for a new topic, but it seems related to this discussion.
Is an alicebot clone, (like the one in my last post) which is 98% of the original,
but uses 0% of the XML schema called AIML, still considered an AIML bot?
|
|
|
|
|
Posted: Mar 11, 2011 |
[ # 15 ]
|
|
Senior member
Total posts: 974
Joined: Oct 21, 2009
|
|
|
|
|