Re: New algorithm for learning languages

Subject: Re: New algorithm for learning languages

From: Matt Arnold <matt.mattarn@gmail.com>

Date: Thu, 1 Sep 2005 13:26:06 -0400

Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:to:subject:in-reply-to:mime-version:content-type:references; b=EgxludHcDUaiw9wtftKwvJpoAC+Vsqq3acVHcqjL62iwz9unW1qch68MK66qAiGXdnJsK1+yxvMDWJ7tXrZBwIlEB4IZ08JhpBLtKlIX8H9lVCly96bmyA2T9vTCIm4qgxroZIYRuwqnuJJPPqTP0r6SL5nxh580FqH0cZbYaaA=

In-reply-to: <e6663d20050901102476e2e086@mail.gmail.com>

References: <e6663d20050901102476e2e086@mail.gmail.com>

Sender: nobody <nobody@digitalkingdom.org>

http://www.scienceblog.com/cms/node/8802

On 9/1/05, Matt Arnold < matt.mattarn@gmail.com> wrote:

     "Science Blog" has an entry today about a new algorithm for a computer to infer the grammatical rules underpinning any language and generate meaningful sentences:
     "In addition to child-directed language, the algorithm has been tested on the full text of the Bible in several languages, on artificial context-free languages with thousands of rules and on musical notation." Is Lojban one of those artificial context-free languages it refers to? If not, perhaps we should contact the researchers and tell them they can test their algorithm on Lojban texts and see if the grammar it infers is the same as the one we have on record.
     I was particularly interested in the fact that this algorithm can generate sentences. I would love to have a program that would output a new Lojban sentence every day for me to translate; and like I said on the beginner's mailing list, I long for the day that jbofi'e will offer grammatically-correct variations on ungrammatical text entered into it.
-epkat

References:

New algorithm for learning languages
- From: Matt Arnold <matt.mattarn@gmail.com>