Newsgroups: comp.ai.nat-lang
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!zombie.ncsc.mil!news.duke.edu!news-feed-1.peachnet.edu!gatech!swrinde!pipex!uknet!festival!edcogsci!usenet
From: poesio@cogsci.ed.ac.uk (Massimo Poesio)
Subject: FRACAS project (Computational Semantics) - Deliverables, WWW Home Page
Message-ID: <POESIO.95Mar10150424@scott.cogsci.ed.ac.uk>
Sender: usenet@cogsci.ed.ac.uk (C News Software)
Nntp-Posting-Host: scott
Organization: University of Edinburgh Cognitive Science
Date: Fri, 10 Mar 1995 15:04:24 GMT
Lines: 94


We would like to announce the World Wide Web page of the FRACAS Project,
at URL:  http://www.cogsci.ed.ac.uk/~fracas/ 

FRACAS - A FRAmework for ComputAtional Semantics - is a two-year
project to study theories of semantic interpretation and their
application in natural language processing, funded by the European
Union (LRE). The participants in the project are CWI Amsterdam, SRI
Cambridge, and the Universities of Edinburgh, Saarbruecken, and
Stuttgart. The deliverables produced in the first year of the project
are available: from our WWW site; by ftp at ftp.cogsci.ed.ac.uk,
directory pub/FRACAS; or by sending mail to fracas@cogsci.ed.ac.uk. (A
list of the deliverables is enclosed below.) Our WWW site includes
pointers to the deliverables, to the WWW pages of the participating
sites, to other projects in computational semantics, and to sites
involved in research in natural language processing. For more
information on the FRACAS Project, contact:

	    The FRACAS Project Administrator
	    University of Edinburgh
	    Centre for Cognitive Science
	    2 Buccleuch Place
	    Edinburgh EH8 9LW, Scotland, UK
	    fracas@cogsci.ed.ac.uk


	        The members of the FRACAS consortium.

--------------------------
Recent FRACAS deliverables
--------------------------
 
Deliverable D7, December 1994: Harmonizing the Approaches

  Our preliminary work towards harmonizing the approaches to semantics that are
studied in FRACAS has led, on the one hand, to the compilation of a list of
basic linguistic phenomena that a semantic theory has to account for;
this list has been used to arrive at the in-depth comparison among the semantic
theories under study presented in Deliverable 8. On the other hand, we
identified a set of basic semantic tools such as generalised quantifiers
theory or abstraction that all of the theories under discussion make use of,
although very often these tools are interpreted in different ways in the
theories under discussion (e.g., although all theories have a notion of
abstraction, the actual properties of the abstraction operation in these
theories differ widely). Both the list of basic linguistic phenomena and the
set of basic semantic tools are discussed in this deliverable.

We also address the issue of whether the problems that are important from a
technical point of view are also important from the point of view of Natural
Language Processing applications, by identifying a set of forms of natural
language use that one could reasonably expect an NLP system will have to deal
with, and by verifying whether the technically challenging data can be
encountered in these forms of text.

-----

Deliverable D8, December 1994: Describing the Approaches

This deliverable contains a detailed discussion of the semantic tools
used by the five semantic theories studied in the FRACAS project -
Discourse Representation Theory, Update and Dynamic Logic, Monotonic
Semantics, Property Theory, and Situation Theory - together with a
presentation of the syntax / semantics interface adopted by each
theory.

-----

Deliverable D9, December 1994: The State of the Art in Computational Semantics:
Evaluating the Descriptive Capabilities of Semantic Theories

In this deliverable we discuss the analyses of the linguistic phenomena
discussed in deliverables D7 proposed by the five semantic theories studied in
the FRACAS project and presented in deliverable D8.

-----

Deliverable D10, January 1994:  Evaluating The State of the Art

      In this deliverable we discuss the themes of the interface to
      semantics, underspecification, contextual reasoning, inference, and
      lexical semantics. We present a brief survey of some implemented
      systems that are based at least in part on some of the approaches to
      semantics that we have described in deliverables D8 and D9. In order
      to ground discussion of the various themes and approaches in this
      deliverable we include an annotated text ("Eurodisney") that
      illustrates the range and variety of semantic phenomena to be found
      even in the simplest newspaper article. We classify the phenomena
      illustrated by this text so as to give some idea of what is within the
      state of the art, and what areas still require a good deal of
      research. The final section amplifies this latter theme, trying to
      summarise the future directions that computational semantics might
      need to take in order  to achieve some of the goals sketched out
      earlier in the document. 

