Good research chance on wikis and blogs

NEW TEXT - Wikis and blogs and other dynamic text sources

Trento, Italy April 3, 2006

newtext@sics.se

http://www.sics.se/jussi/newtext

Call for participation

The EACL 2006 Workshop on New Text will be hosted in conjunction withthe 11th Conference of the European Chapter of the Association for Computational Linguistics ( EACL, http://eacl06.itc.it/ ) that will take place April 3-7, 2006, in Trento, Italy.

New types of text sources, multi-lingual, with numerous cooperating or even adversarial authors and little or no editorial control are one effect of the recently dramatically lowered publication threshold.

Many contain linguistic items or features classically associated with spoken language - combining the high interactivity of dialogue with the low bandwidth of written text and with the multicasting capabilities of digital communication.
New material published today most noticeably includes *blogs* - a genre that has evolved from diaries, logbooks, commentaries, columns, and editorials into a multi-faceted and networked churn of text with widely ranging viewpoints and perspectives and varying application and ambition on the part of the creator. One of the most noticeable charateristics of the blog genre is its opinionated nature and its timeliness. Blog texts are often ill-edited and hastily cobbled together in a language reminiscent of brief notes, spoken asides, or short letters, rather than of essays or newsprint. This, at any rate, is the public perception.

Another emergent genre is that of the *wiki*. More closely patterned on a classic text genre, that of the encyclopedia, wiki texts are written and edited by open teams of authors. In contrast to blogs, wikis have high ambitions as regards factual correctness, persistence, editorial quality, and trustworthiness.
Bridging the two are genres such as discussion boards, web fora, and mailing lists.

Let us call these various new types of text (or indeed other modes of linguistic communication) collectively NEW TEXT.

THIS WORKSHOP is intended to discuss the analysis and application of new text, formulate research measures that are crying out to be taken, discuss which methodological steps are obsoleted, and which babies can be saved from the bath water.

NEW TEXT - Challenge questions

NEW TEXT provides a number of research issues, immediately obvious questions, and tentative applications for our research fields:

1. New possibilities for the philologically inclined: How does new text cast new light on human communicative behaviour? This includes question on style and genre: the characteristics of new text and relations to traditional media. Do blogs in fact resemble spoken language in any important way? Do wikis hold up their promise of qualitative information dissemination?

2. New challenges for building text analysis tools -- how are the today's algorithms portable to new text? This includes questions on multilinguality, code-switching, register variation, and formality melange apparent in new text.

3. New challenges for evaluation methodologies for information access
systems:

+ Can new text, with dynamic information sources and streams of variable quality and impact be plugged into relevance-oriented evaluation frameworks without revising the target notion of text relevance?

+ Some new texts have high social impact; some sink without a trace; some have high import in tightly knit circles and communities. Traditional media have sales figures, citation indices, and distribution analyses. How can the impact of new texts be analyzed?

+ New texts have variable perceived intellectual status and quality -- how can it be measured and predicted?

4. New opportunities for new services -- e.g. linking different types of text in dynamic and interactive sessions of information refinement and elaboration.
Signing up for the workshop

To participate in the workshop: begin by announcing your interest to us (newtext@sics.se) as soon as possible! We may be sending out a data set and a common task for everyone to play with before the workshop.

If you wish to present your work or your ideas at the workshop you are invited to submit full papers on original, unpublished work in the topic area. A presentation should address some of the challenge questions stated above. We are also thinking of making a sample text set available for experimentation for all participants before the workshop.

Submissions should be formatted using the EACL 2006 stylefiles with overt author and affiliation information and not exceeding 8 pages.

The EACL 2006 stylefiles are available at
http://eacl06.itc.it/submission/submission.htm .

LaTeX submissions are much preferred.

Please send your PDF file no later than January 6, 2006, to newtext@sics.se
Each submission will be reviewed at least by two members of the programme committee. Accepted papers will be published in the workshop proceedings.
Dual submissions to the ma in EACL 2006 conference and this workshop are allowed; if you submit to the main session, do indicate this when you submit to the workshop. If your paper is accepted for the main session, you should withdraw your paper from the workshop upon notification by the main session.
Important dates

* Deadline for workshop paper submissions: Jan. 6, 2006
* Notification of workshop paper acceptance: Jan. 27, 2006
* Deadline for camera-ready workshop papers: Feb. 10, 2006

Workshop program committee
* Jussi Karlgren, SICS (chair)
* Shlomo Argamon, IIT
* Bj?rn Gamb?ck, SICS
* Michael Gamon, Microsoft
* Gilad Mishne, University of Amsterdam
* Martin Svensson, SICS
* ?zlem Uzuner, MIT