2022 September status on Danish lexemes in Wikidata

Posted on September 19, 2022 Updated on September 19, 2022

Some statistics:

• 15,047 Danish lexemes in Wikidata compared to 689,114 across all languages. 15th largest language in Wikidata, – Russian has over 100,000 lexemes.
• Links to Danish external identifiers:
  • 7,820 Wikidata lexemes linked to words DanNet and 4,182 indicated as not in DanNet.
  • 1,140 COR lemmas and 213 lexemes indicated as not in COR.
  • 1,103 COR forms and 69 forms indicated as not in COR.
  • 774 Oqaasileriffik online dictionary ID (either Danish, Greenlandic or English)
• 66,618 Danish forms. 18th language, – Estonian has almost 3 million.
• 7,590 Danish senses. 11th language, – Basque has over 30,000 senses.
• 11,960 hyphenation specifications.
• 2,867 unique hyphenation parts: ne, de, te, le, se, ge, en, ning, re, be
• Lexical categories for Danish lexemes:

Interesting individual lexemes:

• rød, – a word with many compounds specified.
• værdipapirfinansieringstransaktionseksponering, longest Danish lexeme attested in Den Europæiske Centralbanks udtalelse af 8. november 2017 om ændringer til EU’s ramme for kapitalkrav til kreditinstitutter og investeringsselskaber
• koronavirus, – a lexeme with many different alternative forms, due to two variations: korona/corona, virusser/virus/vira.
• led (representation), – a word with many homographs

Images from Wikimedia Commons. Licenses available from links at https://ordia.toolforge.org/L2310.

Types:

• Transitive/intransitive/monotransitive/bitransitive verbs
• “Nexual“/”innexual” nouns
• Countable/mass noun
• Compound words
• “Kentaurnominal“
• Color term
• Agent noun
• …

Descriptions, use cases for instance:

• Europarl
• Dansk-russisk ordbog
• Fremmedordbog
• Danish: a comprehensive grammar
• Dansk Grammatik
• Elementær Dansk Grammatik
• Grammatik over det Danske Sprog
• Substantiviske Komplekse Ord Med Subkonfikser I Moderne Dansk
• Solenergianlæg ved Bjerndrup

Alignment with COR

I am working on examining the alignment between

• Missing genitive on nouns and numerals in Wikidata.
• øl (øllen/øllet), plan (planen/planet): They are separate in COR but one lexeme in Den Danske Ordbog and currently in Wikidata.
• Adverbs from adjectives. They are under adjectives in COR.
• Adjectives from verbs (perfektum participium), e.g., ubekymret (in COR as an adjective), bekymret (not in COR as an adjective)
• Various schemes for adjectives, e.g., -sk adjectives do not have differences in grammatical number.

See also the earlier posts Linking from Danish Wikidata lexemes to COR and Part-of-speech tags in Det Centrale Ordregister.

Ordia

Ordia is a web application for Wikidata lexemes.

Tools in the application:

• Lexeme linking with “text to lexemes“
• Language guess with “text to languages“

The tool has apparent around 200 pageviews per day: