A Progress Report on Activities at the Linguistic Data Consortium Benefitting the LREC Community

Proceedings of the 12th Conference on Language Resources and Evaluation (LREC 2020), pages 3449–3456 Marseille, 11–16 May 2020 c⃝ European Language Resources Association (ELRA), licensed under CC-BY-NC 3449 A Progress Report on Activities at the Linguistic Data Consortium Benefitting the LREC Community Christopher Cieri, James Fiumara, Stephanie Strassel, Jon Wright, Denise DiPersio, Mark Liberman Linguistic Data Consortium, University of Pennsylvania 3600 Market Street, Philadelphia, PA 19104 USA {ccieri, jfiumara, strassel, jdwright, dipersio, myl} AT ldc.upenn.edu Abstract This latest in a series of Linguistic Data Consortium (LDC) progress reports to the LREC community does not describe any single language resource, evaluation campaign or technology but sketches the activities, since the last report, of a data center devoted to supporting the work of LREC attendees among other research communities. Specifically, we describe 96 new corpora released in 2018- 2020 to date, a new technology evaluation campaign, ongoing activities to support multiple common task human language technology programs, and innovations to advance the methodology of language data collection and annotation. Keywords: language resources, common task programs, technology evaluations, citizen linguists 1. Introduction Linguistic Data Consortium (LDC) activities include the collection, annotation, processing, distribution, archiving and curation of language resources (LRs) including data, tools and specifications to support language related research, education and technology development worldwide and in any language. LDC also conducts data centered research in a growing number of fields that rely upon language data and engages with the relevant research communities to establish best practices in the application of linguistic analysis and human language technologies to their disciplines. 2. The Consortium Model In 1992 the first goal of the newly formed LDC was to streamline the distribution of LRs, principally data sets. The consortium model developed by the original LDC Advisory Board in 1992 has proven effective and withstood the test of time. Notwithstanding the other successful models of ELRA/ELDA 1 , CLARIN 2 and SADiLaR3, the LDC model persists because it balances between two inevitable but conflicting facts of life: 1) data wants to be free and 2) data creation has its costs and creators need support. Some motivations behind the free data movement surrounding LRs are that broad access to data: encourages the development and spread of technologies that, for example, raise the standard of living; improves the free flow of other sources of information; reinforces the basic human right to use one’s native language; and lowers barriers to participation in the worldwide digital economy thereby reducing the gap between linguistic haves and have-nots. Nevertheless, there are inevitable costs associated with the creation, distribution and maintenance of LRs. These include not only technical costs such as computing cycles, storage, data security and networking bandwidth, but fees and associated legal costs for licensing copyrighted materials, compensation for human subjects and the 1 http://www.elra.info 2 https://www.clarin.eu researchers who manage corpus development and distribution, and the cost of the facilities where they work. LDC’s principal model attempts to find a stable balance among these competing forces by pooling small contributions from many …

Content truncated to 500 words