[95962] in tlhIngan-Hol
[Tlhingan-hol] looking for Klingon corpus for training machine
daemon@ATHENA.MIT.EDU (De'vID)
Fri Apr 5 04:54:19 2013
Date: Fri, 5 Apr 2013 10:53:53 +0200
From: "De'vID" <de.vid.jonpin@gmail.com>
To: tlhIngan-Hol <tlhingan-hol@kli.org>
Errors-To: tlhingan-hol-bounces@stodi.digitalkingdom.org
Are there any good quality Klingon texts which are available
electronically and whose copyright allows them to be used for training
a machine learning algorithm?
I'm looking for both monolingual texts and texts with English
translations. For the former, I think Qov's bologh and {nuq bop bom}
novel have it covered. (Assuming that I can get your permission to use
the text for training a computer to recognise Klingon, Qov?)
For the latter, the KLI's publications are copyright. Furthermore,
they are often not literal translations. This is a good thing for
human readers, but not so good for training a computer. Also, is the
{paq'batlh}'s text available electronically?
Has anyone:
1) trained a machine learning algorithm to identify Klingon text
(i.e., given a text in any language, tell if it's Klingon)
2) attempted to train a machine learning algorithm to translate
Klingon (however badly)?
I'm talking about machine learning/AI algorithms (neural nets and the
like) only, not rule-based systems.
Thanks.
--
De'vID
_______________________________________________
Tlhingan-hol mailing list
Tlhingan-hol@stodi.digitalkingdom.org
http://stodi.digitalkingdom.org/mailman/listinfo/tlhingan-hol
