The efforts on the identification of LRs have been increased considerably at ELRA in the past few years. This has been directly linked to ELRA’s emphasis on identifying already existing resources, wherever they may be developed and stored, easing their availability to their potential users within the Language Technology R&D community and, consequently , helping to reduce the efforts to produce already-available resources.
As a consequence, and further to our Catalogues of ready-for-distribution LRs, intensive work is taking place on the identification of all existing LRs in order to compile them in what we refer to as the Universal Catalogue, which provides all available information for its users.
How do we discover and identify LRs?
The discovery and identification methods can be summarised as follows:
- “LR watch” through ads, announcements in linguistic-related mailing lists: Linguist list, Corpora, LN, ELSNET , etc
- Participation in national, European and international projects related to LR creation.
- Development of expertise on national, European and international research programs.
- Participation in HLT events and conferences.
- Search for information in technical documentation and publications, such as conference proceedings (LREC, Eurospeech, ICSLP , Coling, ACL).
- Information about the existence of LRs by ELDA contacts.
- Mouth-to-ear information.
What is identification upon demand?
Further to this general identification task, ELRA is also carrying out identification upon demand for its members. This latter activity implies searching for specific LRs that our members need and that are not available in our catalogues. Thus, in addition to feeding us with information for our Universal Catalogue, this activity allows us to enrich our own ELRA catalogues and to continue in the forefront of the distribution of LRs for the HL T community. A number of interesting LRs are currently being searched and discussed under this framework. A further positive feature of this modality is that new resources are sometimes created to meet our members’ needs, which helps the field of Language Resources to remain productive and growing.
Please contact Valérie Mapelli for any request on LR identification.