linguistlist.org>
Editor for this issue: Karen Milligan <karenlinguistlist.org>
cs.rutgers.edu>Syntactic and Semantic Complexity in Natural Language Processing Systems Workshop to be held in conjunction with ANLP-NAACL2000 Date: Sunday, April 30 2000 Seattle, Washington. WORKSHOP DESCRIPTION The last decade has seen an explosion in the work done in the development of robust natural language processing systems. A common methodology used in building these systems has been to analyze a sample of the data available (either manually, or automatically for training statistical systems), build statistical/heuristical schemas based upon the analysis, and test the system on a blind sample of the data. Due to this commonly used paradigm, an important area of research that has not been given the attention it deserves is the estimation of syntactic and semantic complexity faced by these systems in the tasks they perform. At the AAAI 1999 Fall Symposium on Question Answering Systems, the problem of semantic complexity, a topic of a 90 minutes panel, motivated a lot of interest and discussion. To continue the investigation of this important issue, in this workshop, we will address the question of complexity as it pertains to the syntax and semantics of natural language. In particular, the workshop will seek to address the following areas: 1) How can we model syntactic and semantic complexity for formal models of natural language? 2) How does complexity impact acquisition of semantic and conceptual information? 3) How does syntactic and semantic complexity impact document classification in information and text retrieval tasks? 4) How do statistical clustering approaches compare to knowledge-based approaches at partitioning and quantifying the semantic space in a document set? 5) Concerning NLP systems that are deployed in the field, how can we quantify the information extraction task and QA task in ways similar to what is currently done with IR tasks and algorithms? 6) How does the estimation of syntactic and semantic complexity impact the evaluation of such systems? 7) Can syntactic and semantic complexity coupled with a history of the past performance of a system be used to predict future performance of the system on a different data set? The workshop invites short papers, full-length papers, proposals for panel discussions, and position statements that deal with any aspect of syntactic and/or semantic complexity of NLP systems. In particular, the workshop is interested in addressing the following topics: - estimation of the syntactic and semantic complexity of specific NLP tasks - semantic complexity and world knowledge - role of syntactic and semantic complexity in system design and testing - syntactic and semantic complexity and its role in the evaluation of NLP systems - use of syntactic and semantic complexity as a performance predictor - relationship between syntactic and semantic complexity FORMAT FOR SUBMISSION Paper submissions should consist of either a short paper (2000 words or less, including references), a position statement (2000 words or less, including references), or a full paper (5000 words or less, including references). Each submission should include a separate title page providing the following information: the title, the type of paper (short/position/full), the word count, a short abstract, names and affiliations of all the authors, the full address of the primary author (or alternate contact person), including phone, fax, and email. Proposals for panels should consist of a short (upto 500 words) description of the proposed panel along with the names of the proposed panelists. Papers and proposals for panel discussions may be submitted by submitting three hard copies or one soft copy (ASCII, or PS) to: Amit Bagga General Electric CRD Room K1-5C38B 1 Research Circle Niskayuna, NY 12309. USA phone: 1-518-387-7077 email: baggaMail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue