November 30th, 2012 | Add a Comment
Held in Stockholm, Sweden, 23 Nov 2012
Download program and presentations here.
I was very happy to attend the workshop “Language in the Digital Age” last week in Stockholm. It was informative and inspring.
The workshop’s venue – Stacken at Nalen’s (a building from the end of XIX century) – is a fascinating example of architectonic re-use. Stacken (literally meaning “The Stack”, but probably a nickname to refer to the boxing ring) was the former boxing gym of the still existing Narva Boxningsklubb. Now Stacken is an cosy conference/banquet room decorated with four thin columns that add status and elegance to events (http://www.cityfinder.se/sv/node/1453#/10)
The speakers and the audience (about 50 people) represented a wide range of interests, from the linguistic needs of the disabled to the requirements underpinning a satisfying digital society. The speakers represented three communities working with language and/or innovation: academics, governmental agencies and entrepreneurs.
The workshop’s working language was Swedish. However, those who wished could speak and interact in English.
I provide this report in English because I think the topics that have been addressed during this workshop are common to many countries and many communities around the world.
From the workshop two main lacks have surfaced:
1) the need for more investments in language technology
2) the necessity of open linguistic resources that both industry and academia can benefit from
Everybody seemed to agree that language technology is one of the crucial area to:
• improve communication,
• increase information sharing,
• streamline processes,
• improve democracy through a special linguistic attention towards disadvantaged groups, such as the visually- or acoustically-impaired and the speakers of minority languages,
• head towards a reliable digital society.
The seminar was open by Lars Borin, professor at Gothenburg University. In his presentation, the speaker showed the most important findings contained in the white paper “The Swedish Language in the Digital Age” (both in Swedish and in English) published by Springer. This white paper is part of a series that promotes knowledge about language technology and its potential. From an investigation carried out among the 30 languages used in the European Union, it turned out that Swedish has currently “fragmented resources” in speech technology, “few or no resources” in machine translation, “fragmented resources” in text analysis, and “moderate resource access”. Lars Borin emphasised that there is indeed much to do and more investments are needed in this field, especially because the use of the Swedish language is larger on the internet than in the physical world. The speaker’s feeling is that in Sweden language is not felt as important for the progress of the society as in Finland or in Norway. And this is certainly something that we all hope will change in future.
Martha D. Brandt from Språkbanken, Gothenburg University – in her talk explained the relations and the specific focus of the different groups forming the META constallation. In particular, the META-NORD group (one of the organizer of this workshop) is responsible for Baltic and Nordic parts of the European Open Linguistic Infrastructure. the META-NORD group has eight countries as members, namely:
Sweden, Latvia, Denmark, Estonia, Norway, Finland, Iceland and Lithuania. META-NORD’s main goal is to “provide a description of the national landscape in terms of language use; language-savvy products and services, language technologies and resources; main actors; public policies and programmes; prevailing standards and practices; current level of development, main drivers and roadblocks; and create this in a simple, clear, standardized format” (http://spraakbanken.gu.se/eng/research/meta-nord).
Rickard Domeij, from Språkrådet (The Swedish Language Council) presented a proposal for a national infrastructure for a digital society. Språkrådet is the “primary regulatory body for the advancement and cultivation of the Swedish language. The council is partially funded by the Swedish government and has semi-official status. The council asserts control over the language through the publication of various books with recommendations in spelling and grammar as well as books on linguistics intended for a general audience, the sales of which are used to fund its operation. The council also works with the five official minority languages in Sweden: Finnish, Meänkieli, Yiddish, Romani and Sami along side with the Swedish Sign Language” (http://en.wikipedia.org/wiki/Swedish_Language_Council). The two main areas of interest for the new infrastructue are:
1) the creation speech-based services that leverage from TV live subtitles, i.e. the conversion of subtitles to spoken speech,
2) proposals and implementation of innovative linguistic resources, the definition of financial priorities and a national strategy.
It was pointed out that a number of special problems arise with some minority languages spoken in Sweden. For instance Sami dialects are quite fragmented and do not have a koine’ (i.e. a supraregional form or lingua franca), the Romani language have graphemes outside the Latin-1charset, and Yiddish has a non Latin-based alphabet. These special aspects require then customized solutions within national language planning.
Mia Ahlgren presented the priorities of The Swedish Disability Federation (Handikappförbunden – HSO), gathering 39 Swedish disability organisations and about 450 000 members. This associations has three top-priorities (see http://www.hso.se/vi-ar-handikappforbunden/In-English/) :
1) A society for all…
The basis for all work at HSO is that everyone is equal and that everybody has the right to decide over their own body and life. The goal is a society for everyone, characterized by solidarity, equality and participation. To achieve this, political initiatives are required in many fields; medical care, support services, education and training, labour market policy, physical planning, culture and information. Almost every political issue has a disability aspect. A main task for the disability movement is to inform about this and to influence decision makers and the general public. One way of exerting such influence has been the developing of Agenda 22.
2) Agenda 22 – a method to realise the UN Standard Rules
Agenda 22 is to realise the UN Standard Rules by drawing up disability policy plans in communities, counties and businesses. The method has been developed by the Swedish disability movement. Guidelines in English are available on how local authorities can work with Agenda 22.
3) Intensive dialogue with the government
Mia Ahlgren’s slides contain a wealth of useful links.
Erik Borälv from Vinnova (VINNOVA- the Swedish Governmental Agency for Innovation Systems – is Sweden’s innovation agency. Their mission is “to promote sustainable growth by improving the conditions for innovations, as well as funding needs-driven research”, http://www.vinnova.se/en/About-VINNOVA/) presented the Open Data project (http://www.vinnova.se/opendata in Swedish). According to a pre-study that has been carried out, open data creates new possibilities for innovation. In this perspective, language plays a main role, since much communication is based on linguistic interaction. As Swedish is a small language, in order to promote Swedish products and high technology in the world Swedish must be understandable to a large audience. It is clear that language technology offers the most suitable solutions in this context, and can boost Sweden’s international visibility.
The three representatives of the needs for language technology for commercial products, namely Svetoslav Marinov from Findwise, Jussi Karlgren from Gavagai and Magnus Merkel from Fodina unanimously pointed out that industry is hungry of advanced open-source linguistic tools and resources such as: lexicons, lemmatizer, POS taggers, syntactic parsers, name-entity recognisers, terminology extractors, together with competent staff, open standards. The common wish is that the same linguistic resources can be shared by industry and academia.
Written by Marina Santini
Computational Linguist, PhD
Filed under: reflections, reports, seminars · Tags: findwise, fodina, gavagai, jussi karlgren, language technology, lars borin, linguistic infrastructure, linguistic resources, magnus merkel, meta-nord, svetoslav marinov
Genres on the Web
Browse with GoogleBooks
Automatic Identification of Genre in Web Pages
Contact me on LinkedIn
- chapter excerpts
- collaborative blogging
- computational models
- reading suggestions
- short reviews
- signed posts
- white papers