© 1985 by British Computer Society
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
On the Selection of a Reduced Set of Indexes

1 Department of Mathematical and Computer Sciences, Michigan Technological University, Houghton, USA, 2 Department of Computer Sciences, National Technical University of Athens, 9 Heroon Polytechnion Ave., Zografou, Athens, Greece
The secondary index selection problem (ISP) is known to be an NP-complete problem. The best possible known algorithm to find an exact solution of the ISP is the one developed by Schkolnick which utilizes the fact that the objective function of the problem satisfies the regularity condition. In a recent paper Ip et al presented a heuristic algorithm which solves the ISP with the added constraint that the creation and storage cost of the optimal selection must not exceed a certain specified cost. The current study observes that in practice this last problem can be translated to the classical ISP but requiring now that no more than k indexes may appear in the optimal set. It is shown that this new problem satisfies the regularity property. The study also presents the modifications needed to Schkolnick's algorithm for solving the problem considered.
* Department of Mathematical and Computer Sciences, Michigan Technological University, Houghton, MI 49931, USA
Department of Computer Sciences, National Technical University of Athens, 9 Heroon Polytechnion Ave., Zografou, Athens(621), Greece