Category: Constraint
Institution: Department of Computer Science, University of
        Arizona
Abstract: This paper introduces a new algorithm for incremental
        attribute evaluation. The algorithm is lazy - it
        evaluates only the attributes that are both affected
        by a change and that are directly or indirectly
        observable by thte user. In this way the wasted work
        of computing values that are never actually used is
        avoided. While the algorithm is not optimal, it will
        perform better than the standard ``optimal''
        algorithm in cases where expensive but optimal
        computations need to be supported. Furthermore, the
        algorithm does not have some of the limitations of
        other algoritmhs. It works for general attributed
        graphs as well as standard attriuted trees. In
        addition, it does not presume any special editing
        model and it supports multiple change points without
        loss of efficiency. Finally, the algorithm can be
        easily extended to a distributed and/or concurrent
        environment. 
Number: TR 89-12
Bibtype: TechReport
Author: Scott E. Hudson
Month: dec
Title: Incremental Attribute Evaluation: A Flexible
        Algorithm for Lazy Update
Year: 1989
Keyword: incremental attribute evaluation, constraint
Address: Tucson, Arizona 85721