Elena Volodina

Language Technology for the multi-lingual society. Nordic conference NoDaLiDa.

The first NoDaLiDa conference - originally called Nordiska DataLingvistikDagarna - was organized in Gothenburg in 1977 where computational linguists could discuss questions relevant to them. Computational linguistics, or Language Technology, can broadly be described as a field of research that automatically handles human language. It covers areas such as speech recognition and speech synthesis, machine translation, effective information search and support of users with special needs. Language Technology also includes creating electronic language resources and infrastructures in support of research in other disciplines, e.g. within digital humanities.

The Nodalida conference series has become a meeting place for researchers within language technology with a focus on the Nordic and Baltic languages. It is organized by the Northern European Association for Language Technology (NEALT) and attracts around 100-150 participants bi-annually. Next Nodalida will be organized in Gothenburg, Sweden, 22-24 May 2017, exactly 40 years after the first Nodalida meeting.

This year we set focus on automatic multilingual analysis of natural languages - an area of research that meets the demands of the modern multi-lingual society, especially in the context of an escalating immigration crisis.

We apply for funding for (a part of) conference centre facilities costs during the three conference days, as well as for subsidizing student participants and volunteers.
Final report

The 40th anniversary of Nodalida, the 21st Nordic Conference on Computational Linguistics, took place in Gothenburg, where it all started, and was a huge success. We had a record number of 184 participants from all over the world taking part in four workshops and the main conference.

Språkbanken at the University of Gothenburg had the honour of organizing the Nodalida 2017 conference, together with the Northern European Association for Language Technology (NEALT), which coordinates the conference series. The conference would allow for researchers in the cross-disciplinary fields of computational linguistics and language technology - with different interests, statuses and seniority - to meet, share their research, and spark discussions.

The participants were researchers, students and industry representatives with an interest in computational linguistics and language technology, from 24 different countries, mainly from the Nordic countries and Europe, but also from further away, such as Hong Kong, Japan, Israel, the United Arab Emirates, and the United States.

The main purpose of the generous funding from RJ was to keep the conference fees low, to be able to attract as many participants as possible, including younger researchers such as master and PhD students, for a lively discourse and to secure the future of the field. The large number of participants shows that we succeeded.

The first conference day had four parallel workshops. One workshop dealt with use of language technology for teaching and learning languages (7 talks and 2 invited speakers), one discussed various aspects of the Universal Dependencies framework (24 talks and 2 invited speakers), and another the Constraint Grammar framework (5 talks) for describing languages, and one dealt with processing historical language (9 talks and 1 invited speaker).

One of the main attractions were the three outstanding keynote speakers, Kyunghyun Cho, Sharon Goldwater, and Rada Mihalcea. They covered questions of multilinguality and natural language analysis from three different viewpoints: processing multilingual big data with neural networks for more effective translation services; automatic speech recognition; and applying language processing for extracting information from the textual information that is generated online in forums and social networks. The talks by the invited speakers were recorded and are now available online for at least 10 years, making them (together with the other conference material) available to a larger audience (see http://nodalida2017.se/).

The main conference had, in addition to the invited speakers, nine sessions, in part parallel, featuring 23 presentations on topics such as syntactic and semantic annotation, working with social media, algorithms for lexical semantics, and machine translation. There were also two poster and demo sessions with  in total 26 contributions. All contributions were double-blind peer-reviewed. The proceedings of the main conference and the workshops were published by both LiUEP (Linköping University Electronic Press) and ACL anthology, both open-access publication channels. Linköping Electronic Conference Proceedings is also ranked as level 1 in the Norwegian list of Scientific Journals, series and publishers. See:

* the main proceedings - http://www.ep.liu.se/ecp/contents.asp?issue=131
* the Workshop on Processing Historical Language: http://www.ep.liu.se/ecp/contents.asp?issue=133
* the Workshop on NLP for Computer Assisted Language Learning and NLP for Research on Language Acquisition: http://www.ep.liu.se/ecp/contents.asp?issue=134
* the Workshop on Universal Dependencies: http://www.ep.liu.se/ecp/contents.asp?issue=135
* the Workshop on Constraint Grammar - Methods, Tools and Applications: http://www.ep.liu.se/ecp/contents.asp?issue=140.

The conference gave multiple possibilities for social gatherings and spontaneous meetings for people with overlapping interests, including a welcome reception hosted by the city of Gothenburg, a conference dinner, lunches, and coffee breaks. This has led to several discussions and collaborations towards future workshops and project proposals, as well as initiated new directions in research. Also, European cooperation in science and technology (COST) networks were present and could attract new members. Nodalida 2017 truly became the meeting place for language technology in the Nordic countries. The long-lasting effects of the conference are yet to be seen, but we expect some important initiatives to mature and get realized later - on par with organisation of ”Language Banks” in Nordic countries (Sweden, Norway, Finland) which emerged as a result of previous Nodalida meetings – a development that has set a long-lasting effect on each country’s language/text-based research in all disciplines – that is, not only within Language Technology, but also within Digital Humanities, Linguistics, Teaching, etc.

Publication list

 
* J Tiedemann, N Tahmasebi (eds). 2017. Proceedings of the 21st Nordic Conference on Computational Linguistics, NoDaLiDa. Sweden, Gothenburg, May 23-24. NEALT Proceedings Series 29. Linköping Electronic Conference Proceedings 131. ISSN (online): 1650-3686. 336 pages (http://www.ep.liu.se/ecp/contents.asp?issue=131)

* G Bouma, Y Adesam (eds). 2017. Proceedings of the NoDaLiDa 2017 Workshop on Processing Historical Language. Sweden, Gothenburg, May 22. NEALT Proceedings Series 32. Linköping Electronic Conference Proceedings 133. ISSN (online): 1650-3740. 58 pages. (http://www.ep.liu.se/ecp/contents.asp?issue=133)

* E Volodina, I Pilán, L Borin, G Grigonyte, K Nilsson Björkenstam (eds). 2017. Proceedings of the Joint 6th Workshop on NLP for Computer Assisted Language Learning and 2nd Workshop on NLP for Research on Language Acquisition at NoDaLiDa, Gothenburg, 22nd May 2017. NEALT Proceedings Series 30, Linköping Electronic Conference Proceedings 134, ISSN (online): 1650-3740. 67 pages. (http://www.ep.liu.se/ecp/contents.asp?issue=134)

* M-C de Marneffe, J Nivre, S Schuster (eds). 2017. Proceedings of the NoDaLiDa 2017 Workshop on Universal Dependencies, 22 May, Gothenburg Sweden. Linköping Electronic Conference Proceedings 135. NEALT Proceedings Series 31. ISSN (online): 1650-3740. 152 pages. (http://www.ep.liu.se/ecp/contents.asp?issue=135)

* E Bick, T Trosterud (eds). 2017. Proceedings of the NoDaLiDa 2017 Workshop on Constraint Grammar - Methods, Tools and Applications, 22 May 2017, Gothenburg, Sweden. NEALT Proceedings Series 33. Linköping Electronic Conference Proceedings 140. ISSN (online): 1650-3686. 31 pages (http://www.ep.liu.se/ecp/contents.asp?issue=140)

Grant administrator
University of Gothenburg
Reference number
F16-1203:1
Amount
SEK 200,000.00
Funding
Research initiation
Subject
Language Technology (Computational Linguistics)
Year
2017