Introduction - ghrgriner/wikidata-lexdata GitHub Wiki

The purpose of this repository is to provide updated translation counts in Wikidata stratified by both source language and target language. Note that here by 'translation' we refer to not just the 'translation' property (P5972) but other properties that can link senses with the same or similar meeting, such as 'item for this sense' (P5137), 'predicate for' (P9970), and 'demonym of' (P6271). The Wikidata documentation [1] states that the 'translation' property is generally unnecessary if the source sense and target sense point to the same item using any of the other three properties listed (subject to distinctions specified in various properties about sense usage, such as semantic gender).

Counts for some of the listed properties have been previously published by Nielsen [2] when lexical data on Wikidata was in its nascency. Since then, Wikidata has grown extensively. Statistics on lexical data in Wikidata are regularly updated on its website and are available on-demand via the Ordia tool [3]. While counts are available stratified by source language on the Wikidata website, counts stratified also by target language are not available. We recently compared published counts of lexemes and senses in Wikidata and Wiktionary (as processed in the DBnary dataset [4-5]) and found that for most languages, the counts in Wikidata were only a small fraction of those in Wiktionary. However, Wikidata did have more senses than Wiktionary for Norwegian, Danish, and Portuguese among the languages we compared. This leads us to wonder whether there are language combinations that also have more counts in Wikidata than in Wiktionary.

For these reasons, we set out to provide updated translation counts stratified by target language in Wikidata (TBD: and compare to Wiktionary / DBnary?). Because we anticipated that we would not be able to calculate the desired counts for all desired language combinations through the Wikidata Query Service, we decided to perform the analysis by downloading the lexicographical data and uploading the data to our own database. We present the results in this repository.

References

[1] Wikidata:Lexicographical data/Documentation/Senses. Retrieved 7 October 2026.

[2] Nielsen, FÅ. Lexemes in Wikidata: 2020 status. Proceedings of the 7th Workshop on Linked Data in Linguistics (LDL-2020), 82-86.

[3] Nielsen, FÅ. Ordia: a web application for Wikidata lexemes. The Semantic Web: ESWC 2019 Satellite Events, 141-146.

[4] Sérasset G (2015), “DBnary: Wiktionary as a Lemon-Based Multilingual Lexical Resource in RDF”. Semantic Web - Interoperability, Usability, Applicability, 6(4):355-361. https://doi.org/10.3233/SW-140147

[5] DBnary (2026). DBnary - Wiktionary as Linguistic Linked Open Data. https://kaiko.getalp.org/about-dbnary/. Retrieved September 14, 2026.