[100002] in tlhIngan-Hol
Re: [Tlhingan-hol] keeping "KLI folklore" words in word lists
daemon@ATHENA.MIT.EDU (De'vID)
Thu Dec 18 05:17:49 2014
In-Reply-To: <CACemAZembpCXjrH8JOtD767Bck=seCzJEnnemdv1Xz5acBtY9Q@mail.gmail.com>
Date: Thu, 18 Dec 2014 11:17:30 +0100
From: "De'vID" <de.vid.jonpin@gmail.com>
To: "tlhingan-hol@kli.org" <tlhingan-hol@kli.org>
Errors-To: tlhingan-hol-bounces@kli.org
On 17 December 2014 at 15:45, Casey Ransberger <casey.obrien.r@gmail.com> wrote:
> I have an idea. Picture a graph where a word has associated symbols marking
> it as significant in different ways, like #canon and #st3, or #noun or #verb
> (I guess I'm describing what's popularly called "tagging.")
This is literally what the source file for the entries in {boQwI'} look like:
<table name="mem">
<!-- Checked on: 2012.11.21. -->
<column name="_id">12920</column>
<column name="entry_name">chu'wI'</column>
<column name="part_of_speech">n:2,deriv,being</column>
<column name="definition">player (of an instrument)</column>
<column name="definition_de">Spieler (eines Musikinstruments)</column>
<column name="synonyms"></column>
<column name="antonyms"></column>
<column name="see_also">{chu'wI':n:1}</column>
<column name="notes"></column>
<column name="hidden_notes"></column>
<column name="components">{chu':v:3}, {-wI':v}</column>
<column name="examples"></column>
<column name="search_tags"></column>
<column name="source">[1] {KGT:src}</column>
</table>
In particular, parts_of_speech is tagged with the categories listed here:
https://code.google.com/p/klingon-assistant/source/browse/KlingonAssistant/data/mem-00-header.xml?r=4841662aed1336886b09142bbc66d7079d8c3273#27
> It would then be pretty trivial to expose via e.g. web UI a way for regular
> folks like us to be able to search and filter based on any of these
> adjectives.
It's trivial in a technical sense, but time-consuming to actually do.
OTOH, the source code for {boQwI'} is publicly available. Feel free to
port it to make a web UI.
Also, the functionality you seem to want is already available from
qurgh's Hol 'ampaS: http://hol.kag.org/ .
> For the CS folks, a further elaboration on this idea might be to find
> interesting relationships between words and start to map those out (either
> automatically, or let users do it wiki-style.) Then it goes past tagging and
> starts to look more like a semantic network.
What sorts of relationships, other than synonyms, antonyms, related
entries, and derivations and components, do you have in mind?
> I got to thinking about this because I found myself really wanting some kind
> of rudimentary syntax highlighting for writing Klingon when I kept saying
> absolutely the wrong thing with typos. That got me thinking about where I
> was going to get the data, which I can't take from the book for copyright
> reasons. If users create the graph, then there's at least an argument that
> this isn't "stolen data."
You're welcome to use the data from {boQwI'} however you like. The
entries were collated mostly by myself, from various sources, with
occasional assistance from a few other people. The German definitions
come from Lieven.
--
De'vID
_______________________________________________
Tlhingan-hol mailing list
Tlhingan-hol@kli.org
http://mail.kli.org/mailman/listinfo/tlhingan-hol
