ScandSum- Summarization network in Scandinavia funded by NORFA , Jan 1, 2002 - Dec 31, 2003

Project participants

KTH-Stockholm
Hercules Dalianis, associate professor, (Docent)
Johan Carlberger
Martin Hassel
Ola Knutsson, fil.lic
Nima Mazdak, email

Bergen University
Koenraad De Smedt, professor
Anja Liseth, email

Cognit
Till Christopher Lech,  email

Centre for Language Technology, Copenhagen
Jürgen Wedekind, Ph.D
Dorte Haltrup
Bart Jongejan

University of Tartu
Kaili Müürisep, email

Abstract
The aim of the network ScandSum is to create and evaluate automatic text summarization for three Scandinavian languages. Swedish, Danish, and Norwegian. Automatic text summarization is the technique where a computer takes an original text and removes the redundant parts and creates a new non-redundant text.
The reason to exploit text summarization techniques are twofold. The amount of text on Internet grows extremely fast and one need tools to filter, extract and to summarize text and the Internet has also gone mobile and one need techniques to push text from WWW to smaller format as WAP, SMS on mobile phones and PDA's.
Today we at NADA-KTH Sweden have developed a text summarization engine available at Internet implemented for five European languages Swedish, English, Spanish, French and German. See SweSum.
By using our general summarization engine and language specific lexicons from each of the three Scandinavians partners it would be very easy to construct text summarization also for Danish (see DanSum) and Norwegian (see NorSum). We expect to do this within a few months. Then we plan to evaluate the summarizers using the same techniques we used when we evaluated the Swedish summarizer.

Norfa Yearbook 2002: Automatic text summarization for the Scandinavian languages. In Holmboe, H. (ed.) Nordisk Sprogteknologi 2002: Årbog for Nordisk Språkteknologisk Forskningsprogram 2000-2004, pp. 153-163. Museum Tusculanums Forlag 2003.

Norfa Yearbook 2003: Porting and evaluation of automatic summarization. In Holmboe, H. (ed.) Nordisk Sprogteknologi 2003. Årbog for Nordisk Språkteknologisk Forskningsprogram 2000-2004 Museum Tusculanums Forlag 2004 (Forthcoming)

Press
Pressrelease Automatisk resumering af avistekster på de skandinaviske sprog. (In Danish). March 26, 2004.
Pressrelease Nu kan du få sammanfattade nyheter i mobilen. (In Swedish), March 26, 2004.
Pressrelease Norfa: DanSum - The first text summariser for Danish, October 9, 2003.
Industry Standard. Nr 3, 21 mars - 18 april 2002. ScandSum tar bort allt strunt: Så blir webben mindre pladdrig, (in Swedish)
Ny Teknik: Sammanfattar internetnyheter så de ryms i mobilfönstret, (in Swedish), 11 okt 2000.
Computer Sweden: Datorförkortade texter gör webben överskådlig, 11 sept, 2000, (in Swedish)

What is automatic text summarization?
What is automatic text summarization?

First ScandSum meeting
Hotel Diplomat, Åre Sweden, April 18-21, 2002 minutes of meeting (photo1, photo2, photo3, photo4)

Second ScandSum meeting
Holmen Fjordhotell, Oslo, June 9, 2002 minutes of meeting

Third ScandSum meeting
Color Hotel Skagen, Denmark Sept 13-15 2002, minutes of meeting (photo1)

Fourth ScandSum meeting
Fefor Høifjellshotell, Fefor, Norway, Jan 25-28, 2003, minutes of meeting (photo1, photo2, photo3)

Fifth ScandSum meeting
Hotell Fjällgården, Åre, Sweden, 4-6 April 2003, minutes of meeting (photo1, photo2, photo3, photo4, photo5)

Sixth ScandSum meeting, UiB, Bergen 7 october 2003, minutes of meeting

Seventh ScandSum meeting,
Hotell Fjällgården, Åre, Sweden, 19-20 April 2004 , minutes of meeting