During the last decade, WordNet has become a powerful resource in (computational) linguistics for various language processing tasks as well as for theoretical research issues. Due to the success of Princeton WordNet, numerous wordnets for further languages have been built or just been started. Therefore, guidelines and principles for comparing existing and acquiring new languages are of utmost importance in the field. Specific lexical properties of new languages should be accounted for. Wordnet developers of less-studied languages can profit from the experience made by the wordnet pioneers, and may also benefit from the feedback provided by wordnet appliers. The workshop will constitute a forum for sharing a common wordnet structure across languages.
The workshop gives emphasis on wordnet structural issues and wordnet standardization rather than implementation issues and will be an excellent forum to bring together researchers that have recently started developing their own wordnets (e.g. Balkans, Scandinavians etc.) or standardizing existing wordnets resources in order to exchange ideas on approaches for linguistic structures of semantic networks and demonstrate their preliminary results to a wider audience. In addition, emphasis will be given on semantic networks for the less-studied languages for which not many such resources are available.
The proposed workshop is entitled "Workshop on Wordnet Structures and Standardization and how these affect Wordnet Applications and Evaluation" and its duration will be one (1) day (morning and afternoon session). The duration of each session will be approximately 5 hours and each session will consist of brief position statements and discussion. One session at the very beginning of the workshop will be used for the keynote speech.
The workshop will emphasize on two major topics: wordnet structures for less-studied languages on the one hand, and wordnet standardization, applications and evaluation on the other. Thus two tracks sessions will be organized within the workshop, each one attempting to highlight different aspects of wordnet development and added value.
One major topic focuses on wordnets for less-studied languages, (i.e. Eastern European, Scandinavian), for which semantic networks have recently started being developed.
The other major topic discusses standardization issues of wordnets with emphasis given on wordnet-related tools as well as on the evaluation of wordnet resources and their applicability in scenarios like IR and sense tagging.
A further goal is to elaborate upon the approaches and discussions presented at the 1st WordNet Conference in Mysore, India.
A preliminary agenda of the workshop as envisaged at this stage will include two separate sessions, one related to structural issues of wordnets for less-studied languages and structures in linguistics in general whereas the second one will concentrate mostly on wordnet standardization and applications as well as on evaluation of wordnets and their implicit structures.
All submissions should describe original or on-going work in the areas of wordnet architecture, linguistic structures, semantic relations and standards for wordnets and wordnet-related tools. All accepted submissions will be published in the Workshop's proceedings. There are no restrictions regarding the background of the participants. Papers focussing on less-studied languages are encouraged.
| Dimitris N. Christodoulakis | Patras University (Greece) |
| Claudia Kunze, Lothar Lemnitzer | University of Tuebingen (Germany) |
| Karel Pala | Masaryk University Brno (Czech Republic) |
| Christiane Fellbaum | Princeton University (USA) |
| Piek Vossen | Irion Technology, Delft (The Netherlands) |
| Kemal Oflazer | Sabanci University Istanbul (Turkey) |
| Jeroen Hoppenbrouwers | Tilburg University (The Netherlands) |
| Randee Tengi | Princeton University (USA) |
| Wim Peters | Sheffield University (GB) |
| Kadri Vider | Universtiy of Tartu (Estonia) |
| Julio Gonzales | UNED Madrid (Spain) |
| Palmira Marrafa | University of Lisboa (Portugal) |
| Paul Buitelaar | DFKI Saarbruecken (Germany) |
| Andreas Wagner | University of Tuebingen (Germany) |
| Erhard Hinrichs | University of Tuebingen (Germany) |
| Simonetta Montemagni | University of Pisa (Germany), t. b. c. |
| R.J.H.M Emers | The Netherlands |
| Deadline for abstract submission | 15th February 2002 |
| Notification of acceptance | 10th March 2002 |
| Final version of paper for workshop proceeding | 5th of April 2002 |
| Workshop | 28th May 2002 |
Wordnets, which are structured along the lines of the Princeton WordNet, have become popular lexical-semantic resources in the field of language technology. Various initiatives to monolingual and multilingual wordnet construction have been launched (EuroWordNet, BalkaNet, Portuguese Wordnet etc.), and numerous language processing tasks rely on wordnet resources and their implicit knowledge structures.
Existing wordnets vary as with respect to their stage of development, coverage of concepts, encoding principles of linguistic contents and semantic relations, and thus their applicability in different NLP tasks. Furthermore, language-specific peculiarities of wordnets have to be considered in the field of cross-lingual applications. Recently attempts have been made towards the construction of wordnets for the less-studied languages, which are in need of reliable standards, yielding at the same time new perspectives on wordnet construction.
This one-day workshop emphasizes two major topics: wordnet structures for less-studied languages on the one hand, and wordnet standardization, evaluation and application on the other hand. The workshop aims at bringing together wordnet builders and wordnet appliers from academia and industries in order to integrate the efforts being made by different sites.
One major topic focuses on wordnets for less-studied languages, i.e. Eastern European and Scandinavian languages which have recently started developing sementic networks in order to exchange new approaches for linguistic structures and architectures of semantic networks and communicate their preliminary results to a wider research community.
The other major topic discusses standardization issues for wordnets and wordnet-related tools, as well as evaluation of wordnet resources and the information encoded in them, and experiences with wordnet applications in the area of information retrieval and sense tagging.
Papers are invited that will describe existing research connected to the topics of the workshop. Each presentation will be 20 minutes long (15 minutes and 5 minutes of discussion). Each submission should indicate: title; author(s); affiliation(s); and contact author's e-mail address, postal address, telephone and fax numbers. Abstracts (maximum 1.500 words, plain-text format) should be sent to the respective contact persons:
Submissions related to Wordnet Structures and Applications for the Less-Studied Languages should be submitted to: dxri@cti.gr
Submissions related to Wordnet Applications, Standardization & Evaluation should be submitted to: kunze@sfs.nphil.uni-tuebingen.de
Accepted papers will be published in the Workshop Proceedings.
The final version of the accepted papers should be no longer than 4,000 words or 10 A4 pages. Instructions for formatting and presentation of the final version will be sent to authors upon notification of acceptance.
Morning Session: Applications, Standardization & Evaluation of Wordnets
Afternoon Session: Wordnet Structures and Applications for Lesser-Studied Languages