IRCS WORKSHOP ON LINGUISTIC DATABASES University of Pennsylvania Philadelphia, USA 11-13 December 2001 http://www.ldc.upenn.edu/annotation/database/ Organized by: Steven Bird, Peter Buneman and Mark Liberman Department of Computer and Information Science, Department of Linguistics, and the Linguistic Data Consortium University of Pennsylvania Funded by the National Science Foundation CALL FOR ABSTRACTS AND PROPOSALS Linguistic databases are digital repositories of structured information intended to document natural language and natural communicative interaction. Over the last decade, linguistic databases have come to stand at the center of empirical research in the language sciences, and in the development of new human language technologies. Like genomic databases, linguistic databases are complex, evolving and richly annotated repositories, and pose interesting challenges for efficient representation, indexing and query. And like most scientific databases, linguistic databases have made little use of standard database technology. The goals of the workshop are to take stock of existing research in linguistic databases, to identify the key problems, and to explore applications of current database research to these problems. More broadly, the workshop will help define the research questions of a new "linguistic database community" and initiate the ongoing interchange of relevant problems and results between this community and the database community at large. The workshop is expected to attract participants from a range of specialties including databases, linguistics, computational linguistics, annotation and markup. There will be tutorial-style presentations on relevant models in each of these areas. The workshop will address a selection of the following topics: MODELS * models for text databases, speech databases, multimodal databases, typological databases, geographical databases (language maps), and metadata repositories * relational, object-oriented and semi-structured models for representing linguistic annotations * representations for specific linguistic datatypes (e.g. databases of aligned parallel text) * modelling temporal and (geo)spatial structure * critical analysis of existing linguistic databases * special problems for systematic data representation posed by linguistic fieldwork LANGUAGES * query of multilayer annotations * linguistic applications/extensions of XML query languages * analysis of existing ad hoc query languages * queries over temporal and (geo)spatial structure OTHER TOPICS * database support (e.g. what standard database technology has proven worthwhile for linguistic databases?) * systematic methods for populating linguistic databases * appropriate indexing methods for linguistic strings and structures * archiving and preservation * metadata standards serving as finding aids for linguistic databases * data provenance / data lineage * annotation servers PROGRAM The program will have a varied format, designed to maximize cross-fertilization among the various specialties, and to allow extended open discussion. Components of the program will include: * tutorials on relevant models from linguistics, databases or annotation, e.g. the structure of lexical entries, semi-structured query languages, models of text and signal annotation * panel sessions on annotated text and lexicons (and possibly others), with position papers and panel discussion, to evaluate competing approaches * full papers reporting new research * demonstrations of systems for creating and/or managing linguistic data TIMETABLE Expressions of interest are welcome anytime, please see the form on the workshop website. If you have any suggestions concerning the workshop, please email the organizers. FRIDAY 14 SEPTEMBER Proposals for tutorials and position papers - please email the organizers FRIDAY 14 SEPTEMBER Abstracts for papers (400 words) and demonstrations (200 words) FRIDAY NOVEMBER 30 Final papers (10 page limit) Registration will be open in September. Please note that participation will be limited by space. PROCEEDINGS The papers will be published in web and hardcopy form (the latter just for workshop attenders). Papers submitted in HTML should be written with the hardcopy version in mind, so a text string which anchors a hyperlink should be directly interpretable, rather than e.g. "visit this link". VENUE The workshop will be held at the Institute for Research in Cognitive Science (IRCS) at the University of Pennsylvania, in Philadelphia, USA. Workshop sessions will take place in IRCS conference rooms, located on the fourth floor of 3401 Walnut Street, adjacent to the university campus, which is two miles west of the city center. The main meeting rooms will be equipped with the usual presentation facilities, including projection and audio facilities. SPONSORSHIP The workshop is being funded by some NSF grants to the University of Pennsylvania. There will be no registration fee, and hotel accomodation will be covered for presenters. USEFUL WEBSITES http://db.cis.upenn.edu Database Research at Penn http://www.ldc.upenn.edu/annotation/ Linguistic Annotation http://www.ldc.upenn.edu/exploration/ Linguistic Exploration http://www.cis.upenn.edu/~ircs/ IRCS homepage http://www.talkbank.org/ NSF TalkBank Project http://www.ldc.upenn.edu/sb/isle.html NSF ISLE Project http://www.language-archives.org/ Open Language Archives Community http://www.upenn.edu/philadelphia/ Philadelphia http://www.facilities.upenn.edu/visitUs/ Getting to Penn ORGANIZERS Steven Bird http://www.ldc.upenn.edu/sb/ Peter Buneman http://www.cis.upenn.edu/~peter/ Mark Liberman http://www.ldc.upenn.edu/myl/ ----