[sword-devel] Strong Numbered TSK Commentary [Re: TSK Commentary into generic cross-references (Peter von Kaehne)]

[Jonathan Morgan]

Strong's numbers may work if you have the numbers in both your source text
and your target text (and I don't think we have many non-English Bibles
tagged), but I'm still not sure that:
a. Inferring them will be straightforward (just as an example, from memory
the coverage of the Hebrew in the KJV OT is not as complete as the Greek in
the NT, and some of the translation links will be one word to many or many
words to one or even many words to many).

b. Going from phrase to phrase will always be a straightforward mapping.  If
both the source text and the target text are comparatively word for word
translations and maintain similar word order, it will probably work
reasonably well.  The less these conditions are true, the less
straightforward the mapping is likely to be.  At least some amount of manual
editing would probably be desirable.

An interesting presentation at BibleTech:2010 was on "Automatic
Concordancing for Scripture in Any Language" by Neil Rees of the British &
Foreign Bible Society.  This was presenting a working attempt to solve a
somewhat similar problem: how to reuse an existing (back of the Bible)
concordance in a different language to make a new concordance for the Bible
in that language.  This includes detecting head words for each concordance
entry and finding the equivalent head word in the target language.  Rather
than using any kind of original language tagging, they had a solution to
gloss directly from the source headword to find the target headword.  I
don't know whether that approach is applicable to this problem, but if it is
it would remove the need to have your target text marked up with Strong's or
to infer the Strong's Numbers from the English headwords.  There was a paper
about it linked from the BibleTech page which provides more detail on the
approach (
http://www.bibletechconference.com/media/2010/NeilRees2010_ReesRidingGlossing_1.pdf).
 I imagine such an approach could have similar problems with matching
phrases rather than words, though I seem to recall that it handled the case
where one word in the source language went to multiple words in the target
language.  In their process, once the basic concordance was produced in the
target language it was then checked and edited until it was ready for
printing.  The demo (going from Swahili to Welsh at the request of the
audience) seemed to work well, though as I know neither Swahili nor Welsh I
can't really vouch for its accuracy.

Jon

On Mon, Apr 26, 2010 at 10:58 PM, Nakamaru Kunio <n_e_o_m at hotmail.com>wrote:
>
> Hello Peter.
>
> Thank you for reply.
> It took a while to work out for language independent TSK.
>
> What I see so far is TSK keywords are based on KJV. so, they
> are "almost" convertable to  strong's numbers with KJV with strong's
> number.
>
> once keywords are in strong's numbers, they become translatable with
> any dictionary to any natural language.
>
> translation can be done runtime or precompiled. or even just leave a
> link to a specified strong dictionary.
>
> Now, does anyone know if strong numbered TSK already exist?
> or knows a better community I should contact?
>
> ybic
>
> Kunio Nakamaru
> _________________________________________________________________
> メール一括チェック！他のWebメールもプロバイダーメールも。
> http://windows7.jp.msn.com/master/hm-popaggre/default.htm
> _______________________________________________
> sword-devel mailing list: sword-devel at crosswire.org
> http://www.crosswire.org/mailman/listinfo/sword-devel
> Instructions to unsubscribe/change your settings at above page
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.crosswire.org/pipermail/sword-devel/attachments/20100427/ee8713de/attachment-0001.html>