Tables for Wiktionary

From Meta, a Wikimedia project coordination wiki

Proposal 1[edit]

Word table
Language table
Meaning table
WordType table
Translation table
Idiom table
Synonym / Antonym  etc table
Pronounciation table
A Wiktionary page has one or more Words
A Word has a Language
A Word has one or more Meanings
A Meaning has a WordType
A Meaning has one or more Translations
A Translation may have a Meaning in another Language, and a definition in the same Language.
Synonym etc has another Meaning in the same Language and an indication what kind of relation exists. 

Other proposals[edit]

Proposal 4[edit]

What it comes down to is:

  • There are Words and Meanings (abstract).
  • Words are represented in Spelling, Pronunciation, etc.
  • Meanings are represented by Descriptions, Multimedia, etc.
  • Words have a Language (English, Latin)
  • Words have some grammatical Classification (noun, verb)
  • Words can have Derivations (good -> goodbye)
  • Words can have Declension (good -> better)
  • Declensions have a grammatical Function (superlative, imperative, plural)

Entities:

  • Spelling: The way an Expression is represented in writing. This is currently the page title.
    • Every Spelling represents one or more Expressions.
    • (Implicit is that the Spelling is in a specific script.)
  • Expression: The uttering that can be spoken, written, represented as a rebus ... This is currently an entry on a page. (Often a single word, sometimes more.)
    • Every Expression has one or more Spellings.
    • An Expression may have one or more Pronunciations. (Actually, in living languages every Expression has at least one, just maybe not in our database.)
    • Every Expression has one or more Usages
    • Every Expression has one Language.
    • An Expression may have a Grammatical Classification.
    • (Internally every Expression will have an identifier to keep them apart.)
  • Usage: The way an expression is included in speech or writing (if you will: an example). This is currently each one of the numbered lines for an entry. Often the same as the Expression itself, in which case some of the dictionaries/pages currently don't write it out explicitly. (In our current implementations this is different from an Expression. This is a trade-off, as otherwise each Usage would currently require a separate page.)
    • Every Usage is for one or more Expressions.
    • Every Usage has one or more Meanings (We could allow the case of no Meaning, especially if there's a declension connecting it to a word which does have a Meaning.)
  • Meaning: What we actually try to convey with our speech or writing. We can't actually include the meaning; it's an abstraction, like Expression. Currently somewhat implicit.
    • Every Meaning has one or more Usages. (When more, they either are synonyms or translations.)
    • Every Meaning has one or more Descriptions. (When more, they ought to be in different languages.)
    • A Meaning can have one or more Sounds/Illustrations/whatever Multimedium .
    • (Internally every Meaning will have an identifier to keep them apart.)
  • Description: A Meaning put into words. Currently, this is what is written on the numbered lines as an explanation.
    • Every Description is for one Meaning.
    • Every Description has one Language (the language of the specific Wiktionary).
  • Language: Speech or writing system. Will surprise us, regardless of how we define it.
    • A language may have one or more Expressions. (All the Expressions of that language treated in the dictionary - Language you look up.)
    • A language may have one or more Descriptions. (All the explanations you find - Language of the Dictionary.)
  • Pronunciation: Pronunciation as represented in some standard system, or as a sound. (For simplicity, sound could be treated as one such standard.)
    • Each pronunciation is for one or more expressions.
    • (Each pronunciation has a Standard and a representation.)
  • Declensions are actually relations, rather than objects:
    • The baseword, the grammatical Function and the declined word.
  • Derivations are actually relations, rather than objects:
    • The baseword and the derived word.
  • Usages in different languages that have the same meaning are translations, for that meaning.
  • Usages in one language that have the same meaning are synonyms for that meaning.
  • Expression may have some other characteristics, depending on the Grammatical Classification.
  • Meanings could be defined as possibly having an opposite, but antonyms are often Expression-based instead.
  • The grammar-based parts are related. They could be defined in several ways.