Newsgroups: comp.constraints,comp.lang.prolog
Path: cantaloupe.srv.cs.cmu.edu!das-news.harvard.edu!news2.near.net!MathWorks.Com!europa.eng.gtefsd.com!howland.reston.ans.net!math.ohio-state.edu!jussieu.fr!univ-lyon1.fr!swidir.switch.ch!newsfeed.ACO.net!Austria.EU.net!EU.net!uunet!news.iij.ad.jp!tyo1-nec!newssv1-nec!dsp.cl.nec!necscc!swl.cl.nec.co.jp!yosikawa
From: yosikawa@swl.cl.nec.co.jp (Masazumi Yoshikawa)
Subject: Re: Timetable construction using CLP?
In-Reply-To: takis@daphne.di.uoa.ariadne-t.gr's message of Tue, 6 Sep 1994 11:15:06 GMT
Message-ID: <YOSIKAWA.94Sep7194934@bols.swl.cl.nec.co.jp>
Sender: news@swl.cl.nec.co.jp
Organization: C&C Research Labs., NEC Corp., Japan
References: <takis.778850106@daphne>
Date: Wed, 7 Sep 1994 10:49:34 GMT
Lines: 36
Xref: glinda.oz.cs.cmu.edu comp.constraints:274 comp.lang.prolog:11219

It's not an application of Constraint Logic Programming, but
you may be interested in my paper:

   Yoshikawa, M.; Kaneko, K.; Nomura, Y.; and Watanabe, M. 1994.
   A Constraint-Based Approach to High-School Timetabling 
   Problems: A Case Study.  In Proceedings of the Twelfth 
   National Conference on Artificial Intelligence (AAAI-94). 

   ABSTRACT: A high-school timetabling problem can be modeled as a 
	  Constraint Relaxation Problem (CRP). Since it is large in scale 
	  and constrained tightly, it is so difficult that several teachers
	  work on for a week or more.
	     This paper describes a general-purpose CRP solver, Coastool, 
	  and its successful experiments on practical high-school 
	  timetabling problems. Coastool has a constraint-based 
	  architecture in which a declarative problem description is 
	  independent of a problem-solving method. A problem can be solved 
	  merely by describing "what is the problem," without describing 
	  "how to solve it." Separating them results in high productivity 
	  for modeling real-world problems. 
	     The effectiveness of this architecture is supported by a 
	  problem-solving method. The method generates a high-quality 
	  initial assignment using Really Full Lookahead Greedy algorithm, 
	  proposed in this paper, and refines it using a hill-climbing 
	  algorithm. Attaching importance to initialization results in 
	  producing a high-quality solution in reasonable time for a large-
	  scale tightly-constrained problem. 
	     This approach has been evaluated by experiments with practical
	  high-school timetabling problems.  As a result, it required only 
	  a few person-days to describe a problem and a few hours to 
	  construct an acceptable timetable. Consequently, Coastool is 
	  efficient as a base of practical applications for high-school 
	  timetabling problems. 
----
MASAZUMI YOSHIKAWA (yosikawa@swl.cl.nec.co.jp)
C&C Research Laboratories, NEC Corporation, JAPAN
