linguistlist.org>
Editor for this issue: Jody Huellmantel <jodylinguistlist.org>
cs.rutgers.edu>*** SECOND CALL FOR PAPERS FOR EMNLP 2001 *** (includes submission instructions; note notification deadline) 2001 Conference on Empirical Methods in Natural Language Processing Sponsored by SIGDAT and the Intelligent Information Systems Institute (IISI). SIGDAT, the Association for Computational Linguistics' special interest group on linguistic data and corpus-based approaches to NLP, invites submissions to EMNLP 2001. The conference will be held at Carnegie Mellon University, Pittsburgh, PA USA on June 3 and 4, immediately preceding the meeting of the North American Chapter of the ACL (NAACL). We are interested in papers from academia, government, and industry on all areas of traditional interest to the SIGDAT community and aligned fields, including but not limited to: * information extraction * information retrieval * language and dialog modeling * lexical acquisition * machine translation * multilingual technologies * question answering * statistical parsing * summarization * tagging * term and named-entity extraction * word sense disambiguation * word, term, and text segmentation Also, to encourage reflection on the current state of the art in corpus-based methods, the conference will have the following theme: "What Works and What Doesn't: Successes and Challenges" Successes --- We solicit papers showing the success of empirical methods in and across application settings. Examples include improvements in information retrieval performance due to employing language modeling techniques; effective use of statistical word segmentation algorithms in machine translation systems; and increased speech recognition accuracy through the incorporation of statistical parsing. Challenges --- It is clear that empirical and corpus-based methods have enjoyed many successes over the past years; but in looking to future accomplishments, the community needs to be aware of the limitations of various techniques and paradigms. We welcome papers that carefully expose and study such limitations. Examples include the identification and exploration of: classes of domains or problems in which popular techniques perform poorly; significant gaps between human and machine performance on tasks where statistical approaches have made great progress; and important practical situations where common assumptions fail to hold. *** We emphasize that we seek submissions that thoughtfully document fundamental limitations, rather than simply report on unsuccessful experiments. *** It is desired that such papers contain thorough examination, via careful experimentation, of the critical factors contributing to the "negative" result. SUBMISSIONS: Requirements: Submissions must describe original, completed, unpublished work, and include concrete evaluation results when appropriate. Papers being submitted to other meetings must provide this information (see submission format); in the event of multiple acceptances, authors are requested to immediately notify the EMNLP program chair (lleeMail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue
cs.rutgers.edu>APOLOGIES FROM MULTIPLE SUBMISSIONS CALL FOR PAPERS CoNLL-2001 Fifth Computational Natural Language Learning Workshop Toulouse, France, July 6-7, 2001 http://lcg-www.uia.ac.be/conll2001/ BACKGROUND CoNLL is the yearly workshop organized by SIGNLL, the Association for Computational Linguistics Special Interest Group on Natural Language Learning (http://www.aclweb.org/signll/). Previous CoNLL meetings were held in Madrid (1997), Sydney (1998), Bergen (1999) and Lisbon (2000). The 2001 event will be held as a two-days workshop at the 39th Annual Meeting of the Association for Computational Linguistics (ACL), July 6-11, 2001 in Toulouse, France. This year, a special theme will be the focus of the workshop: Interaction and Automation in Language Learning Resources Apart from this special theme, the workshop will accept contributions about language learning topics, including, but not limited to: - Computational models of human language acquisition - Computational models of the origins and evolution of language - Machine learning methods applied to natural language processing tasks (speech processing, phonology, morphology, syntax, semantics, discourse processing, language engineering applications) - Symbolic learning methods (Rule Induction and Decision Tree Learning, Lazy Learning, Inductive Logic Programming, Analytical Learning, Transformation-based Error-driven Learning) - Biologically-inspired methods (Neural Networks, Evolutionary Computing) - Statistical methods (Bayesian Learning, HMM, maximum entropy, SNoW, Support Vector Machines) - Reinforcement Learning - Active learning, ensemble methods, meta-learning - Computational Learning Theory analyses of language learning - Empirical and theoretical comparisons of language learning methods - Models of induction and analogy in Linguistics This year's workshop will also accept submissions for a shared task (segmenting a text into clauses-clausing). THE WORKSHOP Main Session Theme: Interaction and Automation in Language Learning Resources The purpose of the special theme is to present and discuss state-of-the-art learning mechanisms for the automated acquisition of language resources (dictionaries, ontologies, grammars) or the automated adaptation of natural language resources/processors to new domains or languages. The dimensions of learning that are of interest for this session include: - The integration of humans/linguists in the learning process - The structure of the training data - The kind of knowledge that is learned - General study of learning methods that are suitable for natural language related tasks Lately there have been new learning mechanisms that use either large amounts of raw data or small sets of carefully constructed tagged training samples. Learning language can be construed as learning numbers or parameters for some statistical or symbolic system, or learning rules assigning structures to input data (or a mix of those). Learning can be done off-line, which introduces the problem of interpreting (if needed) the derived knowledge before its use in an NLP engine; or on-line, which raises user interaction problems. Different approaches are tailored to solve different kinds of problems subject to a different balance of requirements (large vs. small training set, tagged vs. untagged training data, results needs interpretation or can be used as is, etc.). While this session aims at presenting the largest panorama of learning techniques, we encourage submission of work on semi-automated learning techniques that involve interaction with a human during the learning process or the intervention of a linguist for interpreting results. Special Session: Shared Task - Segmenting Text Into Clauses We invite groups to take part in a shared task: Segmenting a Text Into Clauses (Clausing). Participating groups will be provided with the same training and testing material, and will all use the same evaluation criteria, thus allowing comparison between various learning technologies. After Chunking, the CoNLL-2000 shared task, Clausing can be seen as the next step towards a full parsing. More information on this shared task is available at: http://lcg-www.uia.ac.be/conll2001/clauses/ Invited Session: Learning Computational Grammars There will be a special session devoted to the presentation and discussion of results of the EU Learning Computational Grammars project (Coordinator: John Nerbonne). Project participants include: the University of Groningen (The Netherlands, coordinator), University of Antwerp (Belgium), the University of Tuebingen (Germany), SRI Cambridge (UK), the University College Dublin (Ireland), the University of Geneva (Switzerland), and Xerox Grenoble (France). Invited Speaker: Eric Brill SUBMISSIONS Format for Paper Submissions for Main Session Submit an abstract of maximum 1500 words (Postscript or ASCII) by April 6, 2001 electronically to the address below. Authors of accepted abstracts will be invited to produce a full paper to be published in the proceedings of the workshop, which will be available at the workshop for participants, and distributed afterwards by the ACL. Submit main session abstracts to: Walter Daelemans, walter.daelemansMail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue