SYNOPSIS
use Megahal;
Megahal::megahal_initialize();
Megahal::megahal_do_reply(texte, log);
Megahal::megahal_learn_no_reply(texte, log);
Megahal::megahal_cleanup();
ABSTRACT
Megahal is a conversation simulator that learns as you talk to it. It uses a Markov Model to learn how to hold a conversation. It is possible to teach Meagahal to talk about new topics, and in different languages.
DESCRIPTION
- megahal_initialize()
- Initialize megahal brain. Required before other functions are called.
- megahal_do_reply(text, log)
- Update the brain by feeding him the input text. Logging is controlled via the log boolean. Returns the answer of the bot.
- megahal_learn_no_reply()
- Same as above but without making a reply. Much faster and typically used in scripts for quick learning.
- megahal_cleanup()
- You need to call this function to save the brain, and finish Megahal.
FILES
Megahal looks for the following files in the current directory:- megahal.brn
- Megahal brain.
- megahal.ban
- List of words which cannot be used as keywords.
- megahal.log
- Log file for errors.
- megahal.txt
- Log file for the conversation.
- megahal.swp
- Keyword translation for Megahal. If Megahal choose ``why'' as keyword and find a line containing ``why because'' in a line of this file then ``why'' gets translated in ``because'' in the reply.
AUTHOR
This page was written by Laurent Fousse, <[email protected]>.COPYRIGHT AND LICENSE
Copyright 2003 by Jason Hutchens, David N. Welton and others.This library is free software; you can redistribute it and/or modify it under the GNU General Public Licence.
POD ERRORS
Hey! The above document had some coding errors, which are explained below:- Around line 138:
-
You forgot a '=back' before '=head1'