UniLang

Hi,

I'm thinking of the technical side of language dictionary creation. Have you ever tried to create one or heard how those are structured in the database.

I guess simple dictionary needs to have the following:
- two-ways translation for a word
- reference to other words
- several meanings of the word basing on the context,
- several examples of usage of the word
- grammar informations (verb/noun, number, gender, countable or not)
- reference to the root noun/verb form

I'm interested how such thing can be put into the sql database

I'm currently working on such a thing for my constructed language. It's fairly basic at the moment, but it certainly is possible.

I have very extensively thought about this for UniLang and came up with the so-called "Universal Language Resource" (ULR) format, which is used for all UniLang vocabulary lists, phrasebooks, stories, site translations, fill-in-exercises etc.. It's not aimed at dictionaries in particular but they can also be stored in it. The format supports a certain degree of linguistic annotation (gender, case, number, transliteration, usage example, phonetic description etc.. ) and attaching multimedia (audio, imagery, video). The foundation of the format is the idea of multi-way translation between languages. Synonyms and dialectal/scriptural variants are also supported.

This is XML based rather than SQL though. For a real large-size dictionary SQL with a proper search index might be mode adequate.
You can read a bit about ULR here: devdoc.php?page=rescore .

If you're purely interested in dictionaries, you might want to have a look at XDXF instead: http://xdxf.sourceforge.net/ (also XML-based)

This is what I have built: http://nith.4fd.us/d/

Take a look and let me know what you think