Skip to main content

LREC Language Library

After the success of the LRE Map introduced in LREC 2010 – now used in many conferences as a normal step in the submission procedure – FLaReNet and ELRA have launched for LREC 2012 the LREC Language Library.

The LREC Language Library is intended to be a collaborative enterprise of the LRT Community, an important contribution to a “community-built” Open Resource Infrastructure.

An LREC Repository has been prepared, hosting a number of resources on all modalities (speech, text, images, etc.) in as many languages as possible.

When submitting a paper, authors are invited to process some pieces of this data, in the language(s) of choice, in one or more of the possible dimensions addressed by the submission (e.g. POS-tag the data, extract/annotate named entities, annotate temporal information, disambiguate word senses, transcribe audio, etc.).

The processed data are then put back in the Repository, that will become one of the META-SHARE repositories, and will be made available to all the LREC participants before the conference, to be compared and analysed.

This collaborative work on annotation/transcription/extraction/… over the same data (many parallel or comparable) and on a large number of processing dimensions will set the ground for a large Language Library, linked to the LRE Map for the description of the data, where everyone can deposit/create processed data of any sort – all our “knowledge” about language.

The instructions for contributing to the LREC 2012 Language Library can be found here.

The LREC Language Library will be made publicly available as soon as it is ready.