Polynomial search

  • Authors:
  • Ben Shneiderman

  • Affiliations:
  • State University Agricultural and Technical College, Farmingdale, U.S.A.

  • Venue:
  • Software—Practice & Experience
  • Year:
  • 1973

Quantified Score

Hi-index 0.00

Visualization

Abstract

By plotting the key value against the record position in a sorted file and then fitting a least squares polynomial through the points, a fast retrieval technique is determined. The target key of a search is inserted in the polynomial and the first access from the file is made on the basis of the evaluation. Since the maximum deviation can be determined, an efficient local search can be made. If the maximum deviation is less than half the size of the file, polynomial searching is more efficient than binary searching. While this method can be applied in many cases, it is most useful in disk oriented file systems where the goal is to minimize the number of accesses even at the expense of some additional calculation. © 1973 Wiley Periodicals, Inc.