Thomas Krämer


IMTVault: Extracting and Enriching Low-resource Language Interlinear Glossed Text from Grammatical Descriptions and Typological Survey Articles
Sebastian Nordhoff | Thomas Krämer
Proceedings of the 8th Workshop on Linked Data in Linguistics within the 13th Language Resources and Evaluation Conference

Many NLP resources and programs focus on a handful of major languages. But there are thousands of languages with low or no resources available as structured data. This paper shows the extraction of 40k examples with interlinear morpheme translation in 280 different languages from LaTeX-based publications of the open access publisher Language Science Press. These examples are transformed into Linked Data. We use LIGT for modelling and enrich the data with Wikidata and Glottolog. The data is made available as HTML, JSON, JSON-LD and N-quads, and query facilities for humans (Elasticsearch) and machines (API) are provided.