nl-kr-request@cs.rpi.edu (NL-KR Moderator Chris Welty) (08/09/89)
NL-KR Digest (Wed Aug 9 11:11:58 1989) Volume 6 No. 32 Today's Topics: Query - OCR of Arabic Parsing word problems Lexical Functional Grammar, Situations Semantics NLU benchmarking - request for info Formal Semantics Sentence analysis and text linguistics UM90: 2nd Int'l Workshop on User Modeling IJCAI 89 Update Tech Report: Symbol Grounding Problem Submissions: nl-kr@cs.rpi.edu Requests, policy: nl-kr-request@cs.rpi.edu Back issues are available from host archive.cs.rpi.edu [128.213.1.10] in the files nl-kr/Vxx/Nyy (ie nl-kr/V01/N01 for V1#1), mail requests will not be promptly satisfied. If you can't reach `cs.rpi.edu' you may want to use `turing.cs.rpi.edu' instead. ----------------------------------------------------------------- To: nl-kr@cs.rpi.edu Date: 27 Jul 89 14:34:38+0200 >From: wilks yorick <yorick@divsun.unige.ch> Subject: Query - OCR of Arabic Is anyone out there aware of work on the OCR of Arabic letters - -either as research or a commercial product? I would be grateful for information on that or ony any computer-based analysis of Turkish, either parsing, morphology, or redundancy measures on letter sequences. Thanks yorick@nmsu.edu Yorick Wilks Computing Research Lab., NMSU, Las Cruces NM 88003. ------------------------------ To: nl-kr@cs.rpi.edu Date: Mon, 31 Jul 89 10:12:15 EDT >From: "Mark D. LeBlanc" <mdl%unh.edu@RELAY.CS.NET> Subject: Parsing word problems Does anyone know of systems which parse word problems? I'm a PhD student here at UNH and currently on a "literature review wave." Many systems, including the nice arithmetic word problem simulations (Kintsch, Greeno, Cummins et al.) start with the word problem in some propositional form. My question really addresses those systems which do parse sentences/problems into some "internal-meaning" representation. My research area is Intelligent Tutoring Systems, specifically in helping young (K-3) children solve arithmetic word problems. I am aware of the 'original' works of STUDENT (Bobrow), ISAAC (Novak), Bregar,Bayley and Rapp's ATN, SOPHIE, etc. Our current parser is a conceptual analyzer, in the spirit of Schank, Riesbeck, Selfridge and Birnbaum, et al. In short, the question is quite broad including current PC software that you may have seen, tried, or just read about. - Mark LeBlanc mdl@unh.edu ------------------------------ To: nl-kr@cs.rpi.edu >From: unido!unidocv.!kurej@uunet.UU.NET (Petra Kurej) Newsgroups: comp.ai.nlang-know-rep Subject: Lexical Functional Grammar, Situations Semantics Keywords: natural language database front end Date: 1 Aug 89 06:41:00 GMT Reply-To: unidocv!kurej@uunet.UU.NET (Petra Kurej) Hi, we - two students at University of Dortmund - want to design and to implement a natural language database front end. The parser shall be based on the Lexical Functional Grammar formalism. The representation of the functional structure may also comprise aspects of situations semantics. As input language we want to use English and the implementation language shall be PROLOG (esp. qprolog or cprolog). We are therefore interested in information about related topics (literature, experiences, software, etc.). We are looking forward to your responses... . So long, CIAO Antje & Petra - --------------------------------------------------------------------- UUCP: ...mcvax!unido!unidocv!kurej - --------------------------------------------------------------------- ------------------------------ To: nl-kr@cs.rpi.edu Date: Thu, 3 Aug 89 12:36:39 EDT >From: rpg@cs.brown.edu Subject: NLU benchmarking - request for info Reply-To: rpg@cs.brown.edu (Robert Goldman) Some time ago, in this digest, I recall reading about a workshop about evaluation of natural-language understanding programs. Could anyone provide me with more information about this workshop? Was a report produced, or a set of position papers/abstracts? If so, I'd be interested in seeing a copy. Best, Robert Goldman ------------------------------ To: nl-kr@cs.rpi.edu >From: kozma@rex.cs.tulane.edu (John Kozma) Newsgroups: sci.lang,sci.logic,comp.lang.misc,comp.ai.nlang-know-rep Subject: Formal Semantics Keywords: detonational semantics model theory Montague case grammar Date: 6 Aug 89 02:29:27 GMT Reply-To: kozma@rex.cs.tulane.edu (John Kozma) Followup-To: sci.logic Can anyone describe succinctly the distinction (if indeed there is one) between "model theoretic semantics" and "denotational semantics"? Or for that matter, between "denotation" and "extension", or between "intensional" and "extensional" verbs. For anyone familiar with both Goedel's First Incompleteness Theorem and Montague Semantics, which do you consider easier to understand? Finally, can anyone suggest references expressing a combination of concepts from Case Grammar and Montague Semantics? I'm attempting to post this to several different groups, but I'm not sure whether some of them are moderated. Anyway, I will gladly post a summary of responses to the same groups, though I judged it prudent to direct follow-ups to sci.logic. Thanx in advance, John P. Kozma kozma@comus.cs.tulane.edu ------------------------------ To: nl-kr@cs.rpi.edu >From: san@4alley.UUCP (Steve Sanderson) Newsgroups: comp.ai Subject: Sentence analysis and text linguistics Keywords: natural lang sentence text linguistics analysis Date: 24 Jul 89 19:45:19 GMT Organization: Tandem Computers, Austin, TX [[the following article appeared on comp.ai, and was quite garbled by the time it got here, I attempted to reconstruct it as best I could - CW ]] Help! I'm looking for survey information on the different methods of sentence analysis and text linguistics. Since I'm relatively new to this area, I'm interested in working both with individual sentences in isolation *and* with units larger than a sentence, such as a paragraph, etc... The following is some of the information I'm looking for with each method: - Perhaps a brief description of the analysis method or references to descriptions. - Input domain; which domain it was designed for. - What information the analysis yields - Performance characteristics, whether projected or actual. - Any other information you might think was useful for evaluating different sentence analysis methods. Does anyone out there know of where I can obtain this information? I know I can get some introductory books, or some in-depth books but I'm really looking for information to help me evaluate and briefly understand many different methods, then when I find some that fit my requirements, I can delve deeper. Thanks, Steve Sanderson cs.utexas.edu!4alley!san -or- halley!san@cs.utexas.0724f# ------------------------------ To: nl-kr@cs.rpi.edu >From: "Alfred Kobsa" <unido!sbsvax!ak@fb10vax.sbsvax> Date: Tue, 8 Aug 89 11:15:16 +0200 (MET dst) Subject: UM90: 2nd Int'l Workshop on User Modeling Second International Workshop on User Modeling Hawaii, March 30 - April 1, 1990 Call for Participation Objectives User models have recently attracted much research interest in the field of computer dialog systems. It has become evident that the cooperativeness of such systems can be greatly improved if the system has a model of the user available which contains assumptions about his/her background knowledge as well as his/her goals and plans in consulting the system. Research in the field of user models investigates how such assumptions can be automatically created, represented and exploited by the system in the course of an interaction with the user. This workshop (a sequel of one in Germany in 1986) will provide a forum for the discussion of research topics central to the development of user modeling components in dialog systems. The issues to be addressed include (but are not restricted to): -Acquisition of user and student models -Plan recognition -Representation of user models -User stereotypes -Dialog planning and response tailoring -Levels of user expertise -Student modeling and tutoring strategies -Shell systems for user modeling -Conceptual models, mental models An explicit aim of the workshop is to bring together researchers working on user modeling in the fields of Artificial Intelligence and Human-Computer Interaction, as well as researchers working in the related field of student modeling in Intelligent Tutoring Systems. Program chairman David Chin, Univ. of Hawaii at Manoa, 2565 The Mall, + local arrangement Honolulu, HI 96822, U.S.A. Phone: (808) 948-8162 E-mail: chin@cultis.ics.hawaii.edu General Chairman Wolfgang Wahlster, Dept. of Computer Science, Univ. of Saarbruecken and German Research Center for Artificial Intelligence D-6600 Saarbruecken 11, West Germany Phone: +49 681 302 2363 E-mail: wahlster%fb10vax.uni-saarland.dbp.de ...[U.S.: uunet!]unido!sbsvax!wahlster Program committee: Sandra Carberry, Univ. of Delaware, U.S.A. Robin Cohen, Univ. of Waterloo, Canada Thomas Green, Univ. of Cambridge, U.K. Anthony Jameson, Univ. of Nijmegen, Netherlands Aravind Joshi, Univ. of Pennsylvania, U.S.A. Alfred Kobsa, Univ. of Saarbruecken, W. Germany Gordon McCalla, Univ. of Saskatchewan, Canada Kathleen McCoy, Univ. of Delaware, U.S.A. Cecile Paris, Univ. of Southern California, U.S.A. Elaine Rich, MCC, Austin, TX, U.S.A. Derek Sleeman, Univ. of Aberdeen, U.K. Michael Tauber, Univ. of Paderborn, W. Germany To encourage interaction and a broad exchange of ideas, the workshop will be kept to a moderate size - preferably to about 50 participants who will either be presenters or commentators on presentations. Attendance is by invitation only. Interested presenters should send 4 copies of an extended abstract or a full paper of their talk to the program chairman (at least 6 pages, double-spaced). Interested commentators should send him a short statement of interest. Inquiries regarding the program and local arrangements should be directed to David Chin, and all other inquiries to Wolfgang Wahlster. Deadlines: Nov. 31, 1989: Extended abstracts due Dec. 31, 1989: Statement of interest of commentators due Jan. 31, 1990: Notification about acceptance of presentation or attendance. The papers of the workshop may be published in the newly founded international journal 'User Modeling and User-Adapted Interaction'. ------------------------------ To: nl-kr@cs.rpi.edu Date: Wed, 26 Jul 89 15:39:06 EDT >From: dewitt@caen.engin.umich.edu (Kathryn Dewitt) Subject: IJCAI 89 Update CONFERENCE HIGHLIGHTS Invited Speakers: Koichi Furukawa, ICOT will speak Monday, August 21, at 11:10am. The title of his talk is "Fifth Generation Computer Project: Toward a Coherent Framework for Knowledge Information Processing and Parallel Processing". Gerald Edelman, Rockefeller University, will speak Monday August 21, at 2:00pm. The title of his talk is "Neural Darwinism and Selective Recognition Automata". E.D. Dickmanns, Universitat de Bundeswehr Munchen, will speak Wednesday, August 23, at 11:10am. The title of his talk is "Real-Time Machine Vision Exploiting Integrak Spatio-Temporal World Models". Enn Tyugu, Institute of Cybernetics, USSR, will speak Thursday, August 24, at 9:00am. The title of his talk is "Knowledge-Based Programming Environments" Fernado Pereira, AT&T Bell Laboratories, will speak Thursday, August 24, at 11:10am. The title of his talk is "Interpreting Natural Language". Geoffrey Hinton, University of Toronto, will speak Friday, August 25, at 11:10 am. The title of his talk is "Connectionist Learning Procedures". Invited Panels: THE CHALLENGE OF NEURAL DARWINISM - Monday, August 21, 4:15pm. members: Stephen W. Smoliar(chair), Linda Smith, David Zisper, John Holland and George Reeke. ROBOT NAVIGATION - Tuesday, August 22, 9:00am members: David Miller(chair), Rod Brooks, Raja Chatila, Scott Harmon, Stan Rosenschein, Chuck Thorpe, and Chuck Weisbin. HIGH-IMPACT FUTURE RESEARCH DIRECTIONS FOR ARTIFICIAL INTELLIGENCE Tuesday, August 22, 11:10am. members: Perry Thorndyke(Chair), Raj Reddy, and Toshio Yakoi ARTIFICIAL INTELLIGENCE and SPACE EXPLORATION - Tuesday, August 22, 2:00pm members: Peter Friedland(chair), David Atkinson, John Muratore, and Greg Swietek. (HOW) IS AI IMPACTING MANUFACTURING? - Friday, August 25, 9:00am. members: Mark Fox (chair), E.J. van de Kraatz, Dennis O'Connor, and Karl Kempf. ------------------------------ To: nl-kr@cs.rpi.edu >From: harnad@phoenix.Princeton.EDU (S. R. Harnad) Newsgroups: comp.ai Subject: Tech Report: Symbol Grounding Problem Keywords: connectionism language categorization Searle symbol-systems Date: 4 Aug 89 04:37:21 GMT Organization: Princeton University, NJ THE SYMBOL GROUNDING PROBLEM Stevan Harnad Department of Psychology Princeton University ABSTRACT: There has been much discussion recently about the scope and limits of purely symbolic models of the mind and about the proper role of connectionism in cognitive modeling. This paper describes the "symbol grounding problem" for a semantically interpretable symbol system: How can its semantic interpretation be made intrinsic to the symbol system, rather than just parasitic on the meanings in our heads? How can the meanings of the meaningless symbol tokens, manipulated solely on the basis of their (arbitrary) shapes, be grounded in anything but other meaningless symbols? The problem is analogous to trying to learn Chinese from a Chinese/Chinese dictionary alone. A candidate solution is sketched: Symbolic representations must be grounded bottom-up in nonsymbolic representations of two kinds: (1) iconic representations, which are analogs of the proximal sensory projections of distal objects and events, and (2) categorical representations, which are learned and innate feature-detectors that pick out the invariant features of object and event categories from their sensory projections. Elementary symbols are the names of these object and event categories, assigned on the basis of their (nonsymbolic) categorical representations. Higher-order (3) symbolic representations, grounded in these elementary symbols, consist of symbol strings describing category membership relations ("An X is a Y that is Z"). Connectionism is one natural candidate for the mechanism that learns the invariant features underlying categorical representations, thereby connecting names to the proximal projections of the distal objects they stand for. In this way connectionism can be seen as a complementary component in a hybrid nonsymbolic/symbolic model of the mind, rather than a rival to purely symbolic modeling. Such a hybrid model would not have an autonomous symbolic "module," however; the symbolic functions would emerge as an intrinsically "dedicated" symbol system as a consequence of the bottom-up grounding of categories' names in their sensory representations. Symbol manipulation would be governed not just by the arbitrary shapes of the symbol tokens, but by the nonarbitrary shapes of the icons and category invariants in which they are grounded. [Presented at CNLS Conference on Emergent Computation, June 1989 Submitted to Physica D -- Preprint Available] - - Stevan Harnad INTERNET: harnad@confidence.princeton.edu harnad@princeton.edu srh@flash.bellcore.com harnad@elbereth.rutgers.edu harnad@princeton.uucp CSNET: harnad%confidence.princeton.edu@relay.cs.net BITNET: harnad1@umass.bitnet harnad@pucc.bitnet (609)-921-7771 ------------------------------ End of NL-KR Digest *******************