History
Context
In the 1990’s, the funding agencies, in particular the European Commission through the DGXIII[1] Language Engineering sector, headed by Roberto Cencioni, considered that a successful development of language technologies relied significantly on the availability of large-scale Language Resources, elaborated with appropriate standards, methodologies, and data manipulation tools.
They also considered the importance of maximizing the ROI made in the production and packaging of Language Resources to ensure their reusability. Access to extensive annotated spoken and written datasets, along with suitable higher-level Language Resources, would not only directly benefit R&D efforts across various private and public organizations but also promote productive collaboration between academia and industry.
In order to carry out such strategic analysis, the European Commission, and a large number of stakeholders in the language engineering field decided to launch a project called RELATOR.
RELATOR Project
The project sought to establish a comprehensive organizational framework for the generation of Language Resources, catering to both written and spoken language engineering needs. These resources were crucial for the advancement of robust language technology and industry in Europe.
Additionally, the project aimed to assess the viability of establishing a coordinated European network of repositories entrusted with storing, disseminating, and maintaining Language Resources. This initiative aimed to support the overarching objective of making extensive Language Resources readily accessible to European organizations engaged in R&D as well as educational endeavors in the long term.
The primary result of RELATOR was the establishment of ELRA along with the launch of several Language Resources production projects (e.g. SpeechDat family, PAROLE, etc.). This action was conducted through discussions that involved senior representatives of the relevant actors including owners of resources, producers, private and public users, funding bodies, scientific and professional associations. Large industrial companies and prominent R&D laboratories were represented in what was known as the RELATOR Steering Committee, chaired by Parajon-Collada, then Deputy Director General of the DGXIII. The discussions revolved around various aspects of the problematic, encompassing needs and requirements, potential solutions, cooperative approach, and prerequisites for a unified European effort. The Steering Committee’s aim was also to identify, describe, and assess alternative methods and structures at various levels (e.g., organizational, technical, legal) which could ensure the creation, management, and maintenance of a European repository of re-usable Language Resources, and their dissemination to the various types of users.
The project had to present final recommendations for establishing a collaborative infrastructure that would act as a collection, verification, management, and dissemination center, built on the foundation provided by existing European structures and organisations.
The project consortium comprises representatives of major European-wide bodies and associations, most notably ELSNET, ESCA and EACL, plus an industrial steering committee composed of representatives of leading IT companies, publishers, telecom operators and other providers of electronic information services. The action has been carried out in co-operation with relevant European groups and with on-going initiatives such as EAGLES. It also implied an analysis of existing international structures.
Creation of ELRA
A unanimous decision was made to set up an independent, permanent, centralised organisation. Legal experts were asked to assess various scenarios and recommend the most beneficial one. Of all suggestions, an incorporated company, a grouping of economic interests as allowed by the French law, a non-profit association, the steering committee felt that an independent, not-for-profit, membership-driven association was the easiest to implement and the most acceptable from the liability and the legal standpoints. The association statutes were prepared and deposited in Luxembourg on February 1995 by 16 founding members, under the name of the European Language Resources Association (ELRA).
To handle every issue related to the association affairs, ELDA, the Evaluations & Language resources Distribution Agency, was established, as the ELRA operational body. From the beginning, ELDA was entrusted with the development and the execution of ELRA’s strategies and plans, in addition to handling the distribution, and all related issues, of Language Resources.
[1] DGXIII was a Directorate within the European Commission responsible for telecommunications, information market, and innovation, operating from mid-80s to early 90’s. Its new name is Directorate-General for Communications Networks, Content and Technology (DG CONNECT).