We warmly invite researchers to a free workshop on annotating linguistic corpora for low/under-resourced languages.
LexiVault Workshop: Developing Annotated Corpora tools for Under-resourced Languages
Where: Queen Mary University of London, Mile End Campus
When: 10 AM – 4PM, 12–13 June 2025
Join us for a two day hands-on workshop exploring LexiVault, a user-friendly, open-source web tool developed by Samantha Wray, Hind Saddiki and Daisy Li as part of the SAVANT project for querying annotated lexicons, especially those of low-resource languages. Psycholinguistic research on lesser-studied languages often requires researchers to build corpora and compute measures like word frequency and phonological neighborhood density from scratch.
LexiVault closes that gap by making these metrics easily accessible and searchable. Currently, the tool hosts lexicons for Tagalog, Bangla, and multiple Arabic dialects, with searchable annotations including part of speech tags, morpheme frequency, transition probability, and more.
We'd like to expand our offerings while helping you convert your language data to a useable, shareable resource, whether you're starting with raw audio, a text corpus, or existing annotations. This workshop is intended for those with any amount of corpus or behavioral data that they would like to process or annotate further for storage and usage on the LexiVault site.
The focus of this two-day workshop will differ from individual to individual depending on the starting state of your dataset and your interests, but could take the following forms:
- automatic transcription of auditory data to create a text corpus from speech
- stemming a text corpus to create a list of morphemes and their frequencies
- part-of-speech tagging a text corpus
- calculating minimal pairs and phonological neighborhood density from a text corpus.
All paths lead to your resource being in a form you (and others!) can easily query in the future.To book a spot onto the workshop (places are limited), or express your interest in future workshops, click here.
