Randomized competitive algorithms for successful and unsuccessful search on self-adjusting linear lists

Abstract

This paper studies the classical Dictionary problem using a self-adjusting linear list. We design and analyze randomized, on-line algorithms for a sequence of successful and unsuccessful searches which are competitive with off-line algorithms. Our algorithms combine our ps bit technique which speeds up unsuccessful search with the randomized move-to-front scheme of Reingold, Westbrook, and Sleator, which they used to speed up successful search.