P87-1025 This structure is then used in a constraint-propagation process that attempts to assign
P86-1008 at some adjacent junction . The constraint-propagation procedure for Kimmo systems proceeds
P86-1008 automata and the general nature of constraint-propagation methods suggests that a significantly
P86-1008 , it is uncertain whether the constraint-propagation method enjoys an advantage on
P86-1008 this space indi - vidually , the constraint-propagation algorithm summarizes whole sets
P86-1008 COMBINATIONS OF POSSIBILITIES The constraint-propagation algorithm differs from the Kimmo
P86-1008 and ( 2,3,3 ) are excluded . The constraint-propagation algorithm never enumerates the
P86-1008 dictionaries hooked up to the constraint-propagation algorithms , and consequently
P86-1008 reduce the amount of search . The constraint-propagation algorithm for generation has
P86-1008 Like the Kimmo machinery , the constraint-propagation machinery is concerned with the
P01-1028 proposed . ( Brew , 1992 ) combines a constraint-propagation mechanism with a shift-reduce
P86-1008 Limited tests indicate that the constraint-propagation algorithm for Kimmo generation
P86-1008 possible at some position , the constraint-propagation method summarizes the possibilities
P86-1008 best-known early example of a constraint-propagation procedure ( cf. Win - ston ,
P86-1008 implementing a modular , non-searching constraint-propagation algorithm ( see Winston ( 1984
P87-1034 of features would correspond to constraint-propagation nodes , while tuples of sets
P86-1008 ALGORITHM CHARACTERISTICS The constraint-propagation algorithm shares with the Waltz
P86-1008 can still fail to hold if the constraint-propagation method is insufficiently powerful
P86-1008 the hard one , we believe the constraint-propagation system is an appropriate step
P86-1008 al. , 1983 ) is possible for the constraint-propagation method . BIG - MACHINE precompilation
hide detail