LREC

LREC
Name	LREC
Status	Active
Frequency	Biennial/Triennial (varies)
Discipline	Computational linguistics; Natural language processing; Human language technology
First	2000
Country	International

LREC

LREC is an international conference series focused on human language technology, computational linguistics, and language resources. It gathers researchers, engineers, policymakers, librarians, and industry representatives from institutions such as European Language Resources Association, Université Paris Diderot, University of Cambridge, Massachusetts Institute of Technology, and Stanford University to discuss corpora, lexica, speech databases, ontologies, and evaluation frameworks. Participants often include contributors from Google Research, Facebook AI Research, Microsoft Research, Amazon Web Services, and national research centers like CNRS, Max Planck Society, and NICT.

Overview

LREC is a focal point for presentations on resource creation, annotation, distribution, and reuse, bringing together projects associated with Penn Treebank, WordNet, Europarl Corpus, Universal Dependencies, and OpenSubtitles. Typical topics intersect with initiatives led by ELRA, ACL, ISCA, COLING, and EMNLP, and often reference standards promoted by ISO committees and tools from GitHub repositories maintained by teams at Carnegie Mellon University, University of Edinburgh, Johns Hopkins University, and ETH Zurich. The conference emphasizes multilingual datasets used by systems developed at DeepMind, Baidu Research, Alibaba DAMO Academy, and institutions such as Tsinghua University and Peking University.

History and Editions

The conference series originated at the turn of the 21st century with program committees drawing from researchers at University of Southern California, University of Pennsylvania, University of California, Berkeley, and University of Maryland. Over successive editions, keynote speakers have included scholars associated with Noam Chomsky, Yorick Wilks, Geoffrey Hinton, Yoshua Bengio, and practitioners from IBM Watson and Bell Labs. Editions have been hosted in cities such as Lisbon, Malta, Valletta, Istanbul, Reykjavik, Athens, Granada, Marrakesh, and Rome with organizing committees drawn from Università di Pisa, Università di Trento, University of Groningen, Ghent University, and University of Malta. The program has evolved alongside milestones like the rise of statistical machine translation, the introduction of neural networks in NLP, and advances exemplified by models from OpenAI and research groups at Google DeepMind.

Conference Structure and Activities

Typical LREC editions feature plenary sessions, oral presentations, poster sessions, workshops, tutorials, and demos, often coordinated with special sessions organized by groups from ELRA and Linguistic Data Consortium. Workshops have connected researchers behind projects like FLORES, MULTILINGUAL BERT, XLM-R, and evaluation campaigns such as those coordinated by WMT and SemEval. Tutorials have been delivered by teams at University of Oxford, University College London, McGill University, and Australian National University. The demo track showcases systems from startup accelerators, corporate labs including NVIDIA, Intel Labs, Samsung Research, and public institutions such as European Commission research programs and national funding agencies like ERC and NSF.

Research Themes and Impact

Research at the conference spans corpus linguistics exemplified by projects like the British National Corpus, lexicography akin to Longman Dictionary of Contemporary English, speech processing echoing work at Bell Labs Speech Research, and semantics linked to initiatives such as FrameNet and PropBank. Themes include multilingualism, low-resource languages supported by fieldwork from teams at SIL International and Max Planck Institute for Psycholinguistics, evaluation methodology influenced by BLEU and ROUGE metrics from groups at IBM Research and Microsoft Research Redmond, and ethical considerations resonant with policy debates in Council of Europe and UNESCO. Impact is visible in downstream deployments by firms such as Salesforce Research, Alphabet Inc., Baidu, and in standards adoption by ISO/TC 37 and repositories managed by DataCite and Zenodo.

Organization and Sponsorship

Organization typically involves collaboration among academic units, national research institutes, and professional associations including ELRA, ACL Special Interest Groups, and local universities like Università di Roma La Sapienza or Universidade de Lisboa. Sponsorship has come from industrial partners like Amazon, Google, Facebook, Microsoft, and regional funding bodies such as Horizon Europe, national ministries, and private foundations including Wellcome Trust and GATES Foundation. Logistics rely on conference services used by organizers at IEEE and event planners that have supported gatherings for ICML and NeurIPS.

Proceedings and Publications

Proceedings are published as open-access volumes by organizations akin to ELRA and often cataloged in digital libraries maintained by ACL Anthology, Zenodo, LREC Proceedings Series, and national libraries such as Bibliothèque nationale de France or British Library. Papers frequently cite datasets like Tatoeba Project and toolkits including NLTK, spaCy, Stanford CoreNLP, and Moses. Special issues and edited volumes stemming from LREC workshops appear in journals published by Springer Nature and Elsevier, while selected papers join collections indexed by Scopus and Web of Science.

Category:Conferences in computational linguistics